Home https://jhylin.github.io/Data_in_life_blog/ A blog website about using software tools to decipher drug discovery and development data quarto-1.7.31 Thu, 02 Apr 2026 11:00:00 GMT A test drive of Marimo notebook in Quarto blog Jennifer HY Lin https://jhylin.github.io/Data_in_life_blog/posts/23_Marimo/Marimo_test.html This is a Marimo notebook using Marimo v0.14.3 that I’ve started to test in January 2026 inside a Quarto blog.

Few things to note so far after using the new Quarto-Marimo extension:

  1. I’ve found out in Januray that only Marimo version 0.13.3 works and likely not other newer versions (noted a few issues logged about this on the extension’s GitHub repo). However, in February, this may not be the case (see below).

  2. This Quarto-Marimo integration works in a Quarto project such as a blog.

  3. The reactivity of Marimo notebooks inside Quarto blog may take a few seconds up to a minute to load.


Sample code

Below are sample code from Marimo with my little adaptation for a test drive.

Just another Quarto project


When the slider is moved to the left or right, the word, “here!” will change accordingly.


More things

And we can respond!

2026 is here!


Reading a .csv file using pandas

This part is more challenging as I need to go back and forth with different versions of Marimo to see how to load a dataframe.

After testing a few different versions of marimo, I have figured out how to read a simple .csv file using pandas. The versions of marimo that will work are 0.13.3, 0.14.3 and 0.17.3. It may initially show an error message saying no module named “pandas” is found, but after waiting for a minute or two, it’ll load on the static HTML page and a dataframe will appear (as shown below).


According to this clip about Quarto-Marimo extension (especially the last part about Marimo plugin for Quarto), polars dataframe library may not work as well as pandas due to some upstream limitation of polars as not all of its depended software packages support WASM (WebAssembly) yet.

To read a .csv file, I’ve found the best way is to use URL as the data source (likely due to the nature of WASM-enabled HTML page and this reminds me of my old blog post that used shinylive).

Back to top ]]> Marimo Notes https://jhylin.github.io/Data_in_life_blog/posts/23_Marimo/Marimo_test.html Thu, 02 Apr 2026 11:00:00 GMT Using Python in R - reticulate Jennifer HY Lin https://jhylin.github.io/Data_in_life_blog/posts/22_Back_to_basics/2_Using_python_in_r_reticulate.html
Background

The main goal here is to try using Python and R concurrently in a notebook setting. Both programming languages have their weaknesses and strengths, and also differences and similarities. By combining them together, we can leverage R’s strong statistical packages and Python’s well-developed and maintained machine learning packages in some of our scientific research work.

For rpy2 package, I’ve tried it in a separate post. Interestingly, I had to pause in the middle of working on it due to an issue about plotting. I’ve since returned to it after trying out reticulate, and have found a way to generate a plot using ggplot2 in rpy2.

For reticulate package, the concept is reversed where we are trying to call Python in a R environment as it is a R-based package. So I’ve changed the title of this post to reflect this.


Setting up notebook environment

Steps I’ve taken:

Initial preparation - this notebook has been created using a virtual environment from uv, with the associated repository created using the uv project management structure as well (please also see this part regarding virtual environment in R). This step has been done prior to the executions of the following steps.

  1. Install R with version at least 4.0 or above, and Python with version at least 3.7 or above (reference). The versions of R and Python used here are suggested by the old reference post about using rpy2 from 2022. So far when I’ve tried it with reticulate package, it works so I’ll leave them as how they are now.

  2. Install R’s language server. All R packages are to be installed via R console/terminal (not in notebooks, otherwise you may run into some error messages).

```{r}
install.packages("languageserver")
```

  1. Install reticulate package as shown in its GitHub repository.

  2. Install rmarkdown package so we can preview .qmd docs (this step may not be required as this is purely for posting on my blog). Some further set-ups may be required if using VS Code, e.g. go to “Settings” of VS Code -> under “User” -> Text Editor -> Files -> Associations -> add *.Rmd as “Item” and rmd as “Value”

  3. Install R extension for VS Code (may not be required as this is because I’m using VS Code. There is another very similar IDE called Positron created by Posit which may not require this step).

  4. Install shiny package (I was reminded to do this while installing these packages… unsure of its direct relevance to my goal here but it doesn’t hurt so I’ve done it)

Quarto markdown documents (.qmd) and R markdown documents (.rmd) are interchangeable as .qmd are based on .rmd so I’ve done a little trial using reticulate package first in .rmd which worked, and then decided to switch to .qmd so that this piece of work may be published on my blog later. The convenient thing here is that you can literally just change the file name ending with .rmd to .qmd to convert a .rmd file into a .qmd file.


Import libraries
# Show R version used
R.version.string
[1] "R version 4.5.1 (2025-06-13)"
# Import R packages/libraries in R
library(reticulate)

# Import Python packages/libraries in R 
# This step is required to install Python packages in R while using reticulate
# Reference: https://rstudio.github.io/reticulate/articles/python_packages.html

# Import pandas
py_require("pandas")
# below likely needed if calling pandas in r code, but since I'll be using python code, 
# we can just import python packages as usual without following code
#pd <- import("pandas")

# Import plotnine
py_require( "plotnine")
#p9 <- import("plotnine")
# Show Python version used
py_config()
python:         /Users/jenniferlin/Data_in_life_blog/.venv/bin/python
libpython:      /Users/jenniferlin/.local/share/uv/python/cpython-3.12.7-macos-x86_64-none/lib/libpython3.12.dylib
pythonhome:     /Users/jenniferlin/Data_in_life_blog/.venv:/Users/jenniferlin/Data_in_life_blog/.venv
virtualenv:     /Users/jenniferlin/Data_in_life_blog/.venv/bin/activate_this.py
version:        3.12.7 (main, Oct 16 2024, 09:10:10) [Clang 18.1.8 ]
numpy:          /Users/jenniferlin/Data_in_life_blog/.venv/lib/python3.12/site-packages/numpy
numpy_version:  1.26.4

NOTE: Python version was forced by VIRTUAL_ENV

Below is a small test snippet of running Python code in R.

# Code snippet below borrowed from https://www.geeksforgeeks.org/r-language/integrating-python-code-with-r/
# Running Python code in R
py_run_string("x = 10")
py_run_string("y = 5")
py_run_string("z = x + y")

# Access the Python variable in R
z <- py$z  
print(z)
[1] 15

I have a feeling now that it’s easier to use reticulate package for the purpose of combining Python and R in one place.


Plotting in R using ggplot2 and plotnine

Next example is using Python code to read a dataframe using Pandas.

```{python}
## This is the code for read_df.py
# import pandas
import pandas as pd

def reading_df(file):
    # Reading a df (using one of my old cleaned datasets from a while ago)
    df = pd.read_csv(file)
    df = df.rename(columns={"QED Weighted": "QED", "Polar Surface Area": "PSA"})
    
    return df
```


ggplot2

Then I’m using R’s well-known plotting package, ggplot2, to plot a scatter plot as an example shown below.

# Read data using read_df.py
source_python("read_df.py")
df <- reading_df("df_ml.csv")

library(ggplot2)

ggplot(df, aes(PSA, QED, colour = Max_Phase)) + 
    geom_point() + 
    geom_jitter()


plotnine

An equivalent Python plotting library based on ggplot2 is plotnine. The same plot is shown below using plotnine with Python code.

from read_df import reading_df
df_data = reading_df("df_ml.csv")

import plotnine as p9
# from plotnine import *
# from plotnine import ggplot, aes, geom_point

plot = (p9.ggplot(data=df_data) + p9.geom_point(mapping=p9.aes("PSA", "QED", color = "Max_Phase")))

plot.show()

All plots shown in this post using reticulate and the other post using rpy2 should look the same by using the same set of data but different code in two different programming languages (which means I’ve somewhat achieved my small goal here!).


Regarding reproducibility of projects in R

I realise after finishing these two posts about using R and Python together that I should’ve created two separate repositories and virtual environments where one is based on Python and the other based on R to improve the reproducibilities of the work (hmm… currently it may look messy with what I’ve done here, as these two posts are in a Python-based virtual environment). It is possible and most likely recommended to set up a R virtual environment when working on R-related projects. This introductory post about renv will help with this.

Back to top ]]> Notes Python R https://jhylin.github.io/Data_in_life_blog/posts/22_Back_to_basics/2_Using_python_in_r_reticulate.html Sun, 09 Nov 2025 11:00:00 GMT Using R in Python - rpy2 Jennifer HY Lin https://jhylin.github.io/Data_in_life_blog/posts/22_Back_to_basics/1_Using_r_in_python_rpy2.html This is something I’ve always wanted to try, that is using both R and Python in one place, so here it is.

Essentially, I think there are two packages that may do the trick - one is rpy2 and the other is reticulate. I’ll try rpy2 first in this post. For the notes about using reticulate package, it’ll be in a separate post.

There will be two different versions of code below where one is based on using R dataframe in Python and ggplot2, and the other one is based on pandas dataframe in Python with ggplot2. From my little experience using both versions of code, I’d say the second one using pandas and ggplot2 works better than the first (other more experienced users may have different opinions of course… there are always new things to learn for sure).


Setting up notebook environment

Steps:

This notebook has been created using a virtual environment using uv, with the associated repository created using the project structure from uv too. This step has been done prior to the executions of the following steps.

  1. Install R with version at least 4.0 or above, and Python with version at least 3.7 or above (as suggested by this post). The versions of Python and R used in this notebook are printed above the plot shown below.

  2. Install R’s language server as shown below.

```{r}
install.packages("languageserver")
```

  1. Install reticulate package (GitHub repo). It’s a useful R interface to Python and will be used in the other post.

  2. Install rmarkdown package (may not be required as this is really just making sure that this note may be posted on my blog later that uses Quarto).

  3. Install R extension for VS Code (may not be required as this is because I’m using VS Code. There is another very similar IDE called Positron created by Posit which may not require this step).

  4. Install rpy2 package. This notebook uses rpy2 version 3.6.4.

Note: the installations of R-related packages need to be done in the R console in RStudio/Positron IDE or R terminal in VS Code. The R code above are not to be executed in Quarto markdown (.qmd) or RMarkdown (.rmd) documents as this will lead to an error message like this, “trying to use CRAN without setting a mirror…” when trying to preview the .qmd or .rmd.


Using R dataframe in python and ggplot2
import rpy2
#import rpy2 as robjects
## To print HTML in notebooks
# import rpy2.ipython.html
# rpy2.ipython.html.init_printing()

from rpy2.robjects.packages import importr

# Using a local converter
from rpy2.robjects import conversion, default_converter
with conversion.localconverter(default_converter):

    utils = importr('utils')
    dataf = utils.read_csv("df_ml.csv")
    
    #print(dataf)
    print(dataf.colnames)

    ## Visualising with ggplot2 in Python
    import rpy2.robjects.lib.ggplot2 as gp

    p = (gp.ggplot(dataf, gp.aes(x='Polar.Surface.Area', y='QED.Weighted')) + 
        # gp.aes(x='Polar.Surface.Area', y='QED.Weighted') + 
        gp.geom_point())

    ## To see plots in an output cell - this doesn't work yet, it generates an empty .png with error message
    from rpy2.ipython.ggplot import image_png
    image_png(p)

    ## this code seems to work by completing the run, but no plot is shown
    # p.plot()

    ## this code also seems to work but with no plot shown
    # print(p)
Error importing in API mode: ImportError('dlopen(/Users/jenniferlin/Data_in_life_blog/.venv/lib/python3.12/site-packages/_rinterface_cffi_api.abi3.so, 0x0002): symbol not found in flat namespace (_R_BaseEnv)')
Trying to import in ABI mode.
[1] "Max_Phase"          "Polar.Surface.Area" "HBA"               
[4] "HBD"                "X.RO5.Violations"   "QED.Weighted"      
[7] "CX.LogP"            "CX.LogD"            "Heavy.Atoms"       

R callback write-console: In addition:   
R callback write-console: Warning messages:
  
R callback write-console: 1:   
R callback write-console: In grSoftVersion() :  
R callback write-console: 
   
R callback write-console:  unable to load shared object '/Library/Frameworks/R.framework/Resources/modules//R_X11.so':
  dlopen(/Library/Frameworks/R.framework/Resources/modules//R_X11.so, 0x0006): Library not loaded: '/opt/X11/lib/libSM.6.dylib'
  Referenced from: '/Library/Frameworks/R.framework/Versions/4.5-x86_64/Resources/modules/R_X11.so'
  Reason: tried: '/opt/X11/lib/libSM.6.dylib' (no such file), '/usr/local/lib/libSM.6.dylib' (no such file), '/usr/lib/libSM.6.dylib' (no such file)
  
R callback write-console: 2:   
R callback write-console: In cairoVersion() :  
R callback write-console: 
   
R callback write-console:  unable to load shared object '/Library/Frameworks/R.framework/Resources/library/grDevices/libs//cairo.so':
  dlopen(/Library/Frameworks/R.framework/Resources/library/grDevices/libs//cairo.so, 0x0006): Library not loaded: '/opt/X11/lib/libSM.6.dylib'
  Referenced from: '/Library/Frameworks/R.framework/Versions/4.5-x86_64/Resources/library/grDevices/libs/cairo.so'
  Reason: tried: '/opt/X11/lib/libSM.6.dylib' (no such file), '/usr/local/lib/libSM.6.dylib' (no such file), '/usr/lib/libSM.6.dylib' (no such file)
  
R callback write-console: 3:   
R callback write-console: In (function (filename = "Rplot%03d.png", width = 480, height = 480,  :  
R callback write-console: 
   
R callback write-console:  failed to load cairo DLL


Coding issues

The first coding issue encountered is that the code to produce a final output of the dataframe won’t transform into a .qmd preview i.e. html page. There is an error message showing such as, “Conversion rules for rpy2.robjects appear to be missing…” when trying to preview this as a .qmd. One of the solutions I’ve found online (code reference) is to set up a local converter in the code and run all rpy2-related code within this converter (as commented in the code above).

The second issue is that I’ve only managed to run the code to show R dataframe in Python but not the plot… hence you will see the error message generated from the code above, with only the column names shown without a plot there. To resolve this issue, more time is spent on looking for answers online. A mini test has been done in a Jupyter notebook to show a short code to achieve a simple plot (link to Jupyter notebook which is based on this code reference) by calling R magic in Python kernel via rpy2. Once this works, I then manage to change the code to import a sample dataframe and set up to display a simple scatter plot as shown below.


Using pandas and ggplot2
# python version used
import sys
print(sys.version)
3.12.7 (main, Oct 16 2024, 09:10:10) [Clang 18.1.8 ]
# Load R magic
%load_ext rpy2.ipython

import pandas as pd

df = pd.read_csv("df_ml.csv")
df = df.rename(columns={"QED Weighted": "QED", "Polar Surface Area": "PSA"})
df.head()
Max_Phase PSA HBA HBD #RO5 Violations QED CX LogP CX LogD Heavy Atoms
0 0 66.81 4 1 0 0.47 3.94 3.94 32
1 0 62.55 3 1 0 0.93 3.38 3.38 25
2 0 73.86 5 1 2 0.12 9.34 9.34 40
3 0 84.22 4 2 0 0.76 2.01 -0.19 26
4 0 40.46 4 0 0 0.62 4.00 4.00 26 
from rpy2.robjects import conversion, default_converter
with conversion.localconverter(default_converter):

    # R version
    %R print(R.version.string)

    # Importing ggplot2
    %R require(ggplot2)

    # Take name of input variable df and assign it to an R variable of the same name
    %R -i df

    # Plot the df
    %R print(ggplot(df, aes(PSA, QED, colour = Max_Phase)) + geom_point())
[1] "R version 4.5.1 (2025-06-13)"
In addition: Warning message:
In (function (mapping = NULL, data = NULL, stat = "identity", position = "identity",  :
  All aesthetics have length 1, but the data has 5670 rows.
ℹ Please consider using `annotate()` or provide this layer with data containing
  a single row.


Dataframe libraries in R and Python

There may be frequent debates about which programming language to use for data cleaning and manipulations. I’m open to use either whenever it suits really. As far as I’m aware, R has dyplr package (part of the core tidyverse packages) that is equivalent to pandas in Python (depending on personal preference and habits, different users will have different opinions about “equivalence” here). There is Polars as well that’s available to use in Python and R (R Polars). I think there’s also a more conventional base R code available that can be used to wrangle dataframes (e.g. a R package called data.table that is based on base R’s data.frame, it appears that it may be harder to learn than dyplr due to syntax but it’ll be great for large datasets). There is a useful post suggested by R users online that introduces and explains data.table’s differences from dyplr.

Back to top ]]> Notes R Python https://jhylin.github.io/Data_in_life_blog/posts/22_Back_to_basics/1_Using_r_in_python_rpy2.html Sat, 08 Nov 2025 11:00:00 GMT Boosted trees Jennifer HY Lin https://jhylin.github.io/Data_in_life_blog/posts/19_ML2-3_Boosted_trees/1_adaboost_xgb.html
Some introductions

I’ve somehow promised myself to do a tree series on machine learning and glad I’ve made it to the boosted trees part (it took a while…). This is also likely my last post on this topic for now as there are other things I want to explore in the near future. Hopefully this is somewhat useful for anyone who’s new to this.


AdaBoost

Adaptive Boost or AdaBoost has originated from Robert E. Schapire in 1990 (Schapire 1990), (Raschka, Liu, and Mirjalili 2022), and was further introduced in 1996 by Robert Schapire and Yoav Freund at a conference which also led to a publication (Freund and Schapire 1997).

As quoted from scikit-learn, an AdaBoost algorithm is doing this:

…fit a sequence of weak learners (i.e., models that are only slightly better than random guessing, such as small decision trees) on repeatedly modified versions of the data.

Note

Weak learner means an ensemble of very simple base classifiers such as decision tree stumps (Raschka, Liu, and Mirjalili 2022)

During the process of running the algorithm, increased weights are given to the incorrectly predicted samples at each iteration, and less weights are given to the correctly predicted ones. This then forces the AdaBoost models to focus more on the less accurately predicted samples with the aim to improve ensemble performance. The predictions from these iterations are combined to produce a final prediction via a weighted majority vote style, which is a well-known signature of tree models. Overall, AdaBoost algorithm can be used for classification or regression problems. The main difference between bagging and boosting is that boosting only uses random subsets of training samples drawn from the training dataset without any replacements (Raschka, Liu, and Mirjalili 2022). One caveat to note is that AdaBoost tend to overfit training data (high variance).

Parameters to tune:

  • n_estimators - number of weak learners

  • learning_rate - contributions of weak learners in the final combination

  • max_depth - depth of trees

  • min_samples_split - minimum required number of samples to consider a split


Gradient boosted trees

Essentially a similar concept is behind gradient boosted trees where a series of weak learners is trained in order to create a stronger ensemble of models (Raschka, Liu, and Mirjalili 2022). However, some differences between these two types of boosted trees (e.g. AdaBoost and XGBoost) should be noted, and rather than describing them in a paragraph, I’ve summarised them in a table below.

AdaBoost XGBoost
trains weak learners based on errors from previous decision tree stump trains weak learners that are deeper than decision tree stumps with a max depth of 3 to 6 (or max number of leaf nodes from 8 to 64)
uses prediction errors to calculate sample weights and classifier weights uses prediction errors directly to produce the target variable to fit the next tree
uses individual weighting terms for each tree uses a global learning rate for each tree
Differences between XGBoost and AdaBoost (Raschka, Liu, and Mirjalili 2022)

XGBoost or extreme gradient boosting (Chen and Guestrin 2016) is one of the most commonly used open-source packages, originally developed at the University of Washington by T. Chen and C. Guestrin, that uses stochastic gradient boosting to build an ensemble of predictive models.

XGBoost documentation - https://xgboost.readthedocs.io/en/stable/index.html

Main parameters to tune as suggested by (Bruce, Bruce, and Gedeck 2020):

  • subsample - controls fraction of observations that should be sampled at each iteration or a subsample ratio of the training instance (as per XGBoost’s scikit-learn API). This is similar to how a random forest operates but without the sample replacement part

  • eta (in XGBoost) or learning_rate (in scikit-learn wrapper interface for XGBoost) - a shrinkage factor applied to alpha (a factor derived from weighted errors) in the boosting algorithm or it simply may be more easily understood as the boosting learning rate used to prevent overfitting

There are of course a whole bunch of other XGBoost parameters that can be tuned, and in order to keep this post at a reasonable reading length, I won’t go through every single one of them, but see this link as an example parameter set for XGBClassifier().

In scikit-learn, there are also two types of gradient boosted tree methods, GradientBoostingClassifer() and HistGradientBoostingClassifier(), in its sklearn.ensemble module (note: equivalent regressor class also available). One way to choose between them is to check sample size first. GradientBoostingClassifer() class is likely better when there is only a small sample size (e.g. when number of sample is less than 10,000), while HistGradientBoostingClassifier() class is likely better when your sample size is at least 10,000+ or more. The HistGradientBoostingClassifier() is a histogram-based gradient boosting classification tree that is mainly inpired by LightGBM.


A demo

In the example below, I’m only using AdaBoost classifier and XGBoost classifier for now. Please note that the dataset used here is very small and the example is likely not going to reflect real-life use case completely (use with care).


Import libraries
import sys
import pandas as pd
import numpy as np
import chembl_downloader
import rdkit
from rdkit import Chem
import pickle

# Import Scikit_mol
import scikit_mol
## Check and clean SMILES
from scikit_mol.utilities import CheckSmilesSanitazion
## Standardise molecules
from scikit_mol.standardizer import Standardizer
## Import fingerprints & descriptors
from scikit_mol.fingerprints import MorganFingerprintTransformer
from scikit_mol.descriptors import MolecularDescriptorTransformer
## Import smi2mol transformer
from scikit_mol.conversions import SmilesToMolTransformer

# Import scikit-learn
import sklearn
from sklearn.model_selection import train_test_split, GridSearchCV
## Data scaler (variance scaling)
from sklearn.preprocessing import StandardScaler
from sklearn.pipeline import make_pipeline
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import AdaBoostClassifier

# Import xgboost classifier
import xgboost
from xgboost import XGBClassifier

print(f"xgboost version used: {xgboost.__version__}")
print(f"scikit-learn version used: {sklearn.__version__}")
print(f"rdkit version used: {rdkit.__version__}")
print(f"scikit_mol version used: {scikit_mol.__version__}")
xgboost version used: 2.0.3
scikit-learn version used: 1.5.0
rdkit version used: 2024.09.4
scikit_mol version used: 0.4.5
# Show Python version used 
print(sys.version)
3.12.7 (main, Oct 16 2024, 09:10:10) [Clang 18.1.8 ]


Data source

Data source is based on ChEMBL database version 33 (as shown by the file name below, “chembl_d_ache_33”), which was downloaded previously from last post (on random forest classifier) by using chembl_downloader.

from pathlib import Path

# Pick any directory, but make sure it's relative to your home directory
directory = Path.home().joinpath(".data", "blog")
# Create the directory if it doesn't exist
directory.mkdir(exist_ok=True, parents=True)

# Create a file path that corresponds to the previously cached ChEMBL data 
path = directory.joinpath(f"chembl_d_ache_33.tsv")

# alternative way to download latest ChEMBL version
# please see post link - https://jhylin.github.io/Data_in_life_blog/posts/17_ML2-2_Random_forest/2_random_forest_classifier.html#data-retrieval-using-chembl_downloader for details
# note: need to specify late_version = latest() first
# path = directory.joinpath(f"chembl_d_ache_{latest_version}.tsv")

if path.is_file():
    # If the file already exists, load it
    df_ache = pd.read_csv(path, sep=',')
else:
    # If the file doesn't already exist, make the query then cache it
    df_ache = chembl_downloader.query(sql)
    df_ache.to_csv(path, sep=",", index=False)
print(df_ache.shape)
df_ache.head()
(7144, 9)
assay_chembl_id target_type tax_id chembl_id canonical_smiles molecule_chembl_id max_phase standard_type pchembl_value
0 CHEMBL1909212 SINGLE PROTEIN 9606 CHEMBL220 CC/C(=C(/CC)c1ccc(O)cc1)c1ccc(O)cc1 CHEMBL411 4.0 IC50 4.59
1 CHEMBL1003053 SINGLE PROTEIN 9606 CHEMBL220 COc1c2occc2cc2ccc(=O)oc12 CHEMBL416 4.0 IC50 4.27
2 CHEMBL2406149 SINGLE PROTEIN 9606 CHEMBL220 COc1c2occc2cc2ccc(=O)oc12 CHEMBL416 4.0 IC50 6.12
3 CHEMBL1909212 SINGLE PROTEIN 9606 CHEMBL220 CC1(COc2ccc(CC3SC(=O)NC3=O)cc2)CCCCC1 CHEMBL7002 -1.0 IC50 4.82
4 CHEMBL3071788 SINGLE PROTEIN 9606 CHEMBL220 O=c1cc(-c2ccc(O)cc2)oc2cc(O)cc(O)c12 CHEMBL28 NaN IC50 7.92

Exploring max phases of compounds in the dataset.

df_ache.value_counts("max_phase", dropna=False)
max_phase
 NaN    6410
 4.0     618
-1.0      69
 2.0      29
 3.0      10
 1.0       7
 0.5       1
Name: count, dtype: int64


Datasets

The definition of “NaN” assigned to max_phase indicates that compounds labelled as “NaN” or “null” have no evidence of showing they’ve reached clinical trials yet, but I’m still keeping them in the dataset (this can also be excluded depending on project goals).

A max_phase of -1 is assigned to molecules with unknown clinical phase status (ChEMBL reference), for which I’ll drop for this particular experiment.

# Fill "NaNs" as "0" first
df_ache.fillna({"max_phase": 0}, inplace=True)
# Select only mols with max_phase of 0 and above
df_ache = df_ache[(df_ache["max_phase"] >= 0)]
df_ache["max_phase"].describe()
count    7075.000000
mean        0.362898
std         1.138520
min         0.000000
25%         0.000000
50%         0.000000
75%         0.000000
max         4.000000
Name: max_phase, dtype: float64
df_ache.value_counts("max_phase", dropna=False)
max_phase
0.0    6410
4.0     618
2.0      29
3.0      10
1.0       7
0.5       1
Name: count, dtype: int64
# Convert max_phase from float to int for the ease of reading predicted outcomes,
# otherwise it'll look like "4., 2., 4., ..."
df_ache = df_ache.astype({"max_phase": int, "canonical_smiles": "string"})
df_ache.dtypes
assay_chembl_id               object
target_type                   object
tax_id                         int64
chembl_id                     object
canonical_smiles      string[python]
molecule_chembl_id            object
max_phase                      int64
standard_type                 object
pchembl_value                float64
dtype: object

Please note: the only molecule with max_phase of “0.5” was converted into “0” after I’ve converted the datatype of max_phase from float to integer (I’ve left it deliberately like this since this is only a demonstration on using scikit_learn’s pipeline along with Scikit-mol (Bjerrum et al. 2023), but in reality this should be handled with care i.e. don’t discard it as different max phase values have different meanings!). Therefore the following max_phase value count will have 6411 molecules with max_phase “0”, rather than the previous number of 6410.

df_ache.value_counts("max_phase", dropna=False)
max_phase
0    6411
4     618
2      29
3      10
1       7
Name: count, dtype: int64


Model building using scikit-learn’s pipeline

Binary classification has been used in my previous posts, e.g. target as max_phase 4 (re-labelled as “1”) with training set of max_phase “null” re-labelled as “0” with their different RDKit molecular features. This time I’ll be using multi-class classification to predict a training set of molecules containing max_phase of 0, 1, 2, 3 and 4.


Define X and y variables
# A sanity check and view on the original dataset 
df_ache
assay_chembl_id target_type tax_id chembl_id canonical_smiles molecule_chembl_id max_phase standard_type pchembl_value
0 CHEMBL1909212 SINGLE PROTEIN 9606 CHEMBL220 CC/C(=C(/CC)c1ccc(O)cc1)c1ccc(O)cc1 CHEMBL411 4 IC50 4.59
1 CHEMBL1003053 SINGLE PROTEIN 9606 CHEMBL220 COc1c2occc2cc2ccc(=O)oc12 CHEMBL416 4 IC50 4.27
2 CHEMBL2406149 SINGLE PROTEIN 9606 CHEMBL220 COc1c2occc2cc2ccc(=O)oc12 CHEMBL416 4 IC50 6.12
4 CHEMBL3071788 SINGLE PROTEIN 9606 CHEMBL220 O=c1cc(-c2ccc(O)cc2)oc2cc(O)cc(O)c12 CHEMBL28 0 IC50 7.92
5 CHEMBL1119333 SINGLE PROTEIN 9606 CHEMBL220 CNCCC(Oc1ccc(C(F)(F)F)cc1)c1ccccc1 CHEMBL41 4 IC50 6.89
... ... ... ... ... ... ... ... ... ...
7139 CHEMBL5216374 SINGLE PROTEIN 9606 CHEMBL220 CCC(C)(C)NCC(O)c1cc(O)cc(OC(=O)N(C)C)c1 CHEMBL5220560 0 IC50 4.70
7140 CHEMBL5216425 SINGLE PROTEIN 9606 CHEMBL220 CC(C)(C)OC(=O)Nc1ccc(O)c(C(=O)NCCCN2CCCCC2)c1 CHEMBL5220695 0 Ki 6.92
7141 CHEMBL5216408 SINGLE PROTEIN 9606 CHEMBL220 O=C1CCOc2cc(OCCCCCSC(=S)N3CCCCC3)ccc21 CHEMBL5220742 0 IC50 7.00
7142 CHEMBL5218078 SINGLE PROTEIN 9606 CHEMBL220 Cn1ccc2cc(-c3cnc4ccc(C(=O)N5CCCCC5)cc4n3)ccc2c1=O CHEMBL5220884 0 IC50 5.27
7143 CHEMBL5216374 SINGLE PROTEIN 9606 CHEMBL220 COC(=O)c1cc2c(OC(=O)N(C)C)cccc2[n+]([11CH3])c1 CHEMBL5220983 0 IC50 7.70

7075 rows × 9 columns

X = df_ache.canonical_smiles
y = df_ache.max_phase
X.shape
(7075,)
y.shape
(7075,)
print(X)
0                     CC/C(=C(/CC)c1ccc(O)cc1)c1ccc(O)cc1
1                               COc1c2occc2cc2ccc(=O)oc12
2                               COc1c2occc2cc2ccc(=O)oc12
4                    O=c1cc(-c2ccc(O)cc2)oc2cc(O)cc(O)c12
5                      CNCCC(Oc1ccc(C(F)(F)F)cc1)c1ccccc1
                              ...                        
7139              CCC(C)(C)NCC(O)c1cc(O)cc(OC(=O)N(C)C)c1
7140        CC(C)(C)OC(=O)Nc1ccc(O)c(C(=O)NCCCN2CCCCC2)c1
7141               O=C1CCOc2cc(OCCCCCSC(=S)N3CCCCC3)ccc21
7142    Cn1ccc2cc(-c3cnc4ccc(C(=O)N5CCCCC5)cc4n3)ccc2c1=O
7143       COC(=O)c1cc2c(OC(=O)N(C)C)cccc2[n+]([11CH3])c1
Name: canonical_smiles, Length: 7075, dtype: string
print(y)
0       4
1       4
2       4
4       0
5       4
       ..
7139    0
7140    0
7141    0
7142    0
7143    0
Name: max_phase, Length: 7075, dtype: int64


Sanitise SMILES

Update on 30th Jan 2025 - A new safe inference mode feature has been introduced in Scikit-mol recently which will be able to handle invalid SMILES, please visit this demonstration notebook for details (thanks to Esben Jannik Bjerrum for mentioning this in the comment below).

This post is going to focus on Scikit-mol which has a manual way to handle SMILES errors, as shown in the code below. Another useful way to deal with SMILES errors is Molpipeline’s SMILES error handling, with an example shown in one of its notebooks. The main difference from what I can see, even though I haven’t got to use it yet, is that Molpipeline takes into account of all the invalid SMILES by giving each invalid SMILES a “NaN” label in the pipeline process - this maintains the matrix shape and good for tracking down the problematic SMILES (molecules).

checksmi = CheckSmilesSanitazion()
# Checking on SMILES (X) 
X_valid, X_errors = checksmi.sanitize(X)

Ideally, X_valid & y_valid will be used for further work, meaning all the invalids or errors will be removed from the training set.

checksmi.errors
SMILES

No SMILES errors are shown.

X_errors
[]

No outputs are there (no errors detected).

print(X_valid)
['CC/C(=C(/CC)c1ccc(O)cc1)c1ccc(O)cc1', 'COc1c2occc2cc2ccc(=O)oc12', 'COc1c2occc2cc2ccc(=O)oc12', 'O=c1cc(-c2ccc(O)cc2)oc2cc(O)cc(O)c12', 'CNCCC(Oc1ccc(C(F)(F)F)cc1)c1ccccc1', 'CCN(CC)CCCC(C)Nc1c2ccc(Cl)cc2nc2ccc(OC)cc12', 'O=c1c(O)c(-c2ccc(O)c(O)c2)oc2cc(O)cc(O)c12', 'O=C1C=CC(=O)C=C1', 'CCc1c2c(nc3ccc(OC(=O)N4CCC(N5CCCCC5)CC4)cc13)-c1cc3c(c(=O)n1C2)COC(=O)[C@]3(O)CC', 'CCc1c2c(nc3ccc(OC(=O)N4CCC(N5CCCCC5)CC4)cc13)-c1cc3c(c(=O)n1C2)COC(=O)[C@]3(O)CC', 'O=C(c1ccc(OCCN2CCCCC2)cc1)c1c(-c2ccc(O)cc2)sc2cc(O)ccc12', 'Oc1c(Cl)cc(Cl)c(Cl)c1Cc1c(O)c(Cl)cc(Cl)c1Cl', 'Clc1ccc(COC(Cn2ccnc2)c2ccc(Cl)cc2Cl)c(Cl)c1', 'CNC(=O)Oc1ccc2c(c1)OC[C@@H]1CCN(C)[C@H]21', 'CNC(=O)Oc1ccc2c(c1)OC[C@@H]1CCN(C)[C@H]21', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'Nc1c2c(nc3ccccc13)CCCC2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2', 'CNC(=O)Oc1ccc2c(c1)OC[C@@H]1CC[N+](C)(C)[C@H]21', 'CNC(=O)Oc1cccc2c1OC[C@@H]1CCN(C)[C@H]21', 'CNC(=O)Oc1cccc2c1OC[C@@H]1CCN(C)[C@H]21', 'CNC(=O)Oc1cccc2c1OC[C@@H]1CCN(C)[C@H]21', 'CNC(=O)Oc1ccc2c(c1)[C@@H]1[C@@H](CCN1C)CO2', 'CN1CC[C@H]2COc3c(O)cccc3[C@H]21', 'COc1ccc2cc3[n+](cc2c1OC)CCc1cc2c(cc1-3)OCO2.[Cl-]', 'COc1ccc2cc3[n+](cc2c1OC)CCc1cc2c(cc1-3)OCO2.[Cl-]', 'COc1ccc2cc3[n+](cc2c1OC)CCc1cc2c(cc1-3)OCO2.[Cl-]', 'COc1ccc2cc3[n+](cc2c1OC)CCc1cc2c(cc1-3)OCO2.[Cl-]', 'COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(C)Cc1ccccc1OC', 'COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(C)Cc1ccccc1OC', 'COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(C)Cc1ccccc1OC', 'COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(C)Cc1ccccc1OC', 'COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(C)Cc1ccccc1OC', 'COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(C)Cc1ccccc1OC', 'COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(C)Cc1ccccc1OC', 'COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(C)Cc1ccccc1OC', 'COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(C)Cc1ccccc1OC', 'COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(C)Cc1ccccc1OC', 'COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(C)Cc1ccccc1OC', 'COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(C)Cc1ccccc1OC', 'COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(C)Cc1ccccc1OC', 'c1ccc2c(c1)[nH]c1ccncc12', 'COc1ccc2c(c[n+](C)c3c4cc5c(cc4ccc23)OCO5)c1OC', 'CN(C)Cc1ccc(CSCc2ccc(-c3ccc(CSCc4ccc(CN(C)C)o4)cc3)cc2)o1', 'CN(C)Cc1ccc(CSCCNc2cc(NCCSCc3ccc(CN(C)C)o3)c([N+](=O)[O-])cc2[N+](=O)[O-])o1', 'CN(C)Cc1ccc(CSCCCCCCCCCSCc2ccc(CN(C)C)o2)o1', 'CN(C)Cc1ccc(CSCCCCCCCCCCSCc2ccc(CN(C)C)o2)o1', 'CN(C)Cc1ccc(CSCc2ccc(CSCc3ccc(CN(C)C)o3)cc2)o1', 'CN(C)Cc1ccc(CSCc2cccc(CSCc3ccc(CN(C)C)o3)c2)o1', 'CN(C)Cc1ccc(CSCc2ccccc2CSCc2ccc(CN(C)C)o2)o1', 'CN(C)Cc1ccc(CSCCCCCCSCc2ccc(CN(C)C)o2)o1', 'CN(C)Cc1ccc(CSCCCCCCCCSCc2ccc(CN(C)C)o2)o1', 'CN(C)Cc1ccc(CSCCNC(=NC#N)NCCSCc2ccc(CN(C)C)o2)o1', 'CN(C)Cc1ccc(CSCCN/C=C(/NCCSCc2ccc(CN(C)C)o2)[N+](=O)[O-])o1', 'CN(C)Cc1ccc(CSCc2ccc(CN(C)C)o2)o1', 'CN(C)Cc1ccc(CSCCSCc2ccc(CN(C)C)o2)o1', 'CN(C)Cc1ccc(CSCCCCSCc2ccc(CN(C)C)o2)o1', 'CN(C)Cc1ccc(CSCCCCCCCSCc2ccc(CN(C)C)o2)o1', 'CN(C)Cc1ccc(CSCCCCCSCc2ccc(CN(C)C)o2)o1', 'CN(C)Cc1ccc(CSCCNC(=O)CSCc2ccc(CN(C)C)o2)o1', 'CN(C)Cc1ccc(CSCCCSCc2ccc(CN(C)C)o2)o1', 'CN(C)Cc1ccc(CSCCNC(=O)NCCSCc2ccc(CN(C)C)o2)o1', 'COc1ccc(C)cc1OC', 'COc1ccc(C)cc1OC', 'CCN(CC)CCNC(=O)c1c(C)[nH]c(/C=C2\\C(=O)Nc3ccc(F)cc32)c1C', 'Clc1ccccc1C(c1ccccc1)(c1ccccc1)n1ccnc1', 'CN(C)C(=O)Oc1cccc([N+](C)(C)C)c1', 'CN(C)CCCC1(c2ccc(F)cc2)OCc2cc(C#N)ccc21', 'O=C(c1ccc(OCCCc2c[nH]cn2)cc1)C1CC1', 'O=C(c1ccc(OCCCc2c[nH]cn2)cc1)C1CC1', 'c1ccc2c(NCCc3ccc(OCCCN4CCCCC4)cc3)c3c(nc2c1)CCCC3', 'Cn1c(=O)c2c(ncn2C)n(C)c1=O', 'Cc1ccccc1', 'Cc1ccccc1', 'Cc1ccc(O)cc1', 'Cc1ccc(O)cc1', 'COc1ccccc1CNCCCCCCNCCSSCCNCCCCCCNCc1ccccc1OC', 'COc1ccccc1CNCCCCCCNCCSSCCNCCCCCCNCc1ccccc1OC', 'COc1ccccc1CNCCCCCCNCCSSCCNCCCCCCNCc1ccccc1OC', 'COc1ccc(Cc2nccc3cc(OC)c(OC)cc23)cc1OC', 'CN1C[C@H](CNC(=O)OCc2ccccc2)C[C@@H]2c3cccc4c3c(cn4C)C[C@H]21', 'CCN(CC)CCS/C(=N\\O)c1nc(C)no1', 'CCN(CC)CCS/C(=N\\O)c1nc(-c2ccccc2)no1', 'CCc1nn(CCCN2CCN(c3cccc(Cl)c3)CC2)c(=O)n1CCOc1ccccc1', 'CC(=O)CCCCn1c(=O)c2c(ncn2C)n(C)c1=O', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1', 'CCN(CC)CCNC(=O)c1ccc(N)cc1', 'CCN(CC)CCNC(=O)c1ccc(N)cc1', 'CCN(CC)CCNC(=O)c1ccc(N)cc1', 'CCN(CC)CCNC(=O)c1ccc(N)cc1', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'CC(C)OC(=O)C(C)(C)Oc1ccc(C(=O)c2ccc(Cl)cc2)cc1', 'c1ccc(CN2CCC(CCNc3ccc(-c4ccccc4)nn3)CC2)cc1', 'c1ccc(-c2ccc(NCCCCN3CCc4ccccc4C3)nn2)cc1', 'c1ccc(-c2ccc(NCCN3CCCCC3)nn2)cc1', 'C[C@]12CC[C@H](O)[C@@](C)(CO)C1C[C@H](O)[C@@]1(C)Oc3cc(-c4ccncc4)oc(=O)c3C(=O)C21', 'Cc1cc(-c2ccccc2)nnc1NCCN1CCOCC1', 'COc1ccc(-c2cc3c(c(=O)o2)C(=O)C2[C@@]4(C)CC[C@H](O)[C@@](C)(CO)C4C[C@H](O)[C@@]2(C)O3)cc1OC', 'C[C@]12CC[C@H](O)[C@@](C)(CO)C1C[C@H](O)[C@@]1(C)Oc3cc(-c4cccnc4)oc(=O)c3C(=O)C21', 'C[C@]12CC[C@H](O)[C@@](C)(CO)C1C[C@H](O)[C@@]1(C)Oc3cc(-c4ccccc4)oc(=O)c3C(=O)C21', 'COc1cc(-c2cc3c(c(=O)o2)C[C@]2(O)[C@@]4(C)C(=O)C=CC(C)(C)[C@]4(O)CC[C@@]2(C)O3)cc(OC)c1OC', 'COc1cc(-c2cc3c(c(=O)o2)C[C@]2(O)[C@@]4(C)C(=O)C=CC(C)(C)[C@]4(O)CC[C@@]2(C)O3)cc(OC)c1OC', 'COc1cc(-c2cc3c(c(=O)o2)C[C@]2(O)[C@@]4(C)C(=O)C=CC(C)(C)[C@]4(O)CC[C@@]2(C)O3)cc(OC)c1OC', 'COc1cc(-c2cc3c(c(=O)o2)C[C@]2(O)[C@@]4(C)C(=O)C=CC(C)(C)[C@]4(O)CC[C@@]2(C)O3)cc(OC)c1OC', 'C[C@]12CC[C@H](O)[C@@](C)(CO)C1C[C@H](O)[C@@]1(C)Oc3cc(-c4ccc(Cl)nc4)oc(=O)c3C(=O)C21', 'COc1cc2nc(N(C)CCCNC(=O)C3CCCO3)nc(N)c2cc1OC', 'COc1ccc(-c2cc3c(c(=O)o2)C[C@]2(O)[C@@]4(C)C(=O)C=CC(C)(C)[C@]4(O)CC[C@@]2(C)O3)cc1OC', 'COc1ccc(-c2cc3c(c(=O)o2)C[C@]2(O)[C@@]4(C)C(=O)C=CC(C)(C)[C@]4(O)CC[C@@]2(C)O3)cc1OC', 'COc1ccc(-c2cc3c(c(=O)o2)C[C@]2(O)[C@@]4(C)C(=O)C=CC(C)(C)[C@]4(O)CC[C@@]2(C)O3)cc1OC', 'COc1ccc(-c2cc3c(c(=O)o2)C[C@]2(O)[C@@]4(C)C(=O)C=CC(C)(C)[C@]4(O)CC[C@@]2(C)O3)cc1OC', 'COc1ccc(-c2cc3c(c(=O)o2)C[C@]2(O)[C@@]4(C)C(=O)C=CC(C)(C)[C@]4(O)CC[C@@]2(C)O3)cc1OC', 'COc1ccc(-c2cc3c(c(=O)o2)C[C@]2(O)[C@@]4(C)C(=O)C=CC(C)(C)[C@]4(O)CC[C@@]2(C)O3)cc1', 'COc1ccc(-c2cc3c(c(=O)o2)C[C@]2(O)[C@@]4(C)C(=O)C=CC(C)(C)[C@]4(O)CC[C@@]2(C)O3)cc1', 'COc1ccc(-c2cc3c(c(=O)o2)C(=O)C2[C@@]4(C)CC[C@H](O)[C@@](C)(CO)C4C[C@H](O)[C@@]2(C)O3)cc1', 'Cc1cccc(C)c1', 'Cc1cccc(C)c1', 'CN(CCn1cc[n+](C)c1/C=N/O)S(C)(=O)=O.[Cl-]', 'C[n+]1ccn(COCCS(C)(=O)=O)c1/C=N/O.[Cl-]', 'C[n+]1ccn(COCC(C)(C)[N+](=O)[O-])c1/C=N/O.[Cl-]', 'C[n+]1ccccc1/C=N/O.[Cl-]', 'C[n+]1ccccc1/C=N/O.[Cl-]', 'C[n+]1ccccc1/C=N/O.[Cl-]', 'C[n+]1ccccc1/C=N/O.[Cl-]', 'C[n+]1ccccc1/C=N/O.[Cl-]', 'C[n+]1ccccc1/C=N/O.[Cl-]', 'C[n+]1ccccc1/C=N/O.[Cl-]', 'C[n+]1ccccc1/C=N/O.[Cl-]', 'C[n+]1ccccc1/C=N/O.[Cl-]', 'C[n+]1ccccc1/C=N/O.[Cl-]', 'C[n+]1ccccc1/C=N/O.[Cl-]', 'C[n+]1ccccc1/C=N/O.[Cl-]', 'C[n+]1ccccc1/C=N/O.[Cl-]', 'COc1ccccc1CNCCCCCCNCCCCCCCCNCCCCCCNCc1ccccc1OC', 'COc1ccccc1CNCCCCCCNCCCCCCCCNCCCCCCNCc1ccccc1OC', 'COc1ccccc1CNCCCCCCNCCCCCCCCNCCCCCCNCc1ccccc1OC', 'O=c1c(O)c(-c2ccc(O)c(O)c2)oc2cc(O)ccc12', 'Cc1ccc(C)cc1', 'Cc1ccc(C)cc1', 'c1ccc2c(c1)cc1ccc3cccc4ccc2c1c34', 'Cn1cc[n+](COCCCc2ccccc2)c1/C=N/O.[Cl-]', 'NC(=O)c1cc[n+](COC[n+]2ccccc2/C=N/O)cc1.[Cl-].[Cl-]', 'NC(=O)c1cc[n+](COC[n+]2ccccc2/C=N/O)cc1.[Cl-].[Cl-]', 'NC(=O)c1cc[n+](COC[n+]2ccccc2/C=N/O)cc1.[Cl-].[Cl-]', 'NC(=O)c1cc[n+](COC[n+]2ccccc2/C=N/O)cc1.[Cl-].[Cl-]', 'NC(=O)c1cc[n+](COC[n+]2ccccc2/C=N/O)cc1.[Cl-].[Cl-]', 'NC(=O)c1cc[n+](COC[n+]2ccccc2/C=N/O)cc1.[Cl-].[Cl-]', 'NC(=O)c1cc[n+](COC[n+]2ccccc2/C=N/O)cc1.[Cl-].[Cl-]', 'NC(=O)c1cc[n+](COC[n+]2ccccc2/C=N/O)cc1.[Cl-].[Cl-]', 'CCN1CC[C@@]2(C)c3cc(OC(=O)NC(C)(C)C)ccc3CC12', 'CCCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(Cc3ccccc3)C1C2', 'CCCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CCC)C1C2', 'CC(C)OC[n+]1ccn(C)c1/C=N/O.[Cl-]', 'CCN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3CC12', 'CCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CC)C1C2', 'CCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CC)C1C2', 'CCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CC)C1C2', 'CCCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CC)C1C2', 'CCCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CC)C1C2', 'CCCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CC)C1C2', 'CCCCCCNC(=O)Oc1ccc2c(c1)[C@@H]1CCN(CCC)C1C2', 'CCCCCCNC(=O)Oc1ccc2c(c1)[C@@H]1CCN(CCC)C1C2', 'CCCCNC(=O)Oc1ccc2c(c1)[C@@H]1CCN(CC)C1C2', 'CCCCCCNC(=O)Oc1ccc2c(c1)[C@@H]1CCN(CC)C1C2', 'CCN1CC[C@@]2(C)c3cc(OC(=O)NCc4ccccc4)ccc3CC12', 'CCN1CC[C@@]2(C)c3cc(OC(=O)N[C@@H](C)c4ccccc4)ccc3CC12', 'CCCCCCCNC(=O)Oc1ccc2c(c1)[C@@H]1CCN(CCC)C1C2', 'CCN1CC[C@H]2c3cc(OC(=O)NC)ccc3CC21', 'CCCNC(=O)Oc1ccc2c(c1)[C@@H]1CCN(CC)C1C2', 'Cn1cc[n+](COCc2cccc3ccccc23)c1/C=N/O.[Cl-]', 'Cn1cc[n+](COCc2ccccc2)c1/C=N/O.[Cl-]', 'CCCCOC[n+]1ccn(C)c1/C=N/O.[Cl-]', 'Cn1cc[n+](COCC(C)(C)C)c1/C=N/O.[Cl-]', 'CCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CC)C1C2', 'CCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CC)C1C2', 'CCCCCCCNC(=O)Oc1ccc2c(c1)[C@@H]1CCN(CC)C1C2', 'CCN1CC[C@@]2(C)c3cc(OC(=O)NC)ccc3CC12', 'CCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1C2', 'CCCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1C2', 'O=c1c(O)c(-c2ccc(O)cc2)oc2cc(O)cc(O)c12', 'CC(OC[n+]1ccn(C)c1/C=N/O)C(C)(C)C.[Cl-]', 'CCCCCCCCOC[n+]1ccn(C)c1/C=N/O.[Cl-]', 'O=c1cc(-c2ccc(O)c(O)c2)oc2cc(O)cc(O)c12', 'c1ccc2c(NCCCCCCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCCCCCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCCCCCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCCCCCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCCCCCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCCCCCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCCCCCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCCCCCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'Clc1cccc(Cl)c1CO/N=C/c1cc[n+](CCCc2ccccc2)cc1', 'CC12CC3CC(C)(C1)CC(N)(C3)C2', 'Clc1ccc(COC(Cn2ccnc2)c2ccc(Cl)cc2Cl)cc1', 'CN(C)C(=O)Oc1ccc[n+](C)c1.[Br-]', 'CN(C)C(=O)Oc1ccc[n+](C)c1.[Br-]', 'CN(C)C(=O)Oc1ccc[n+](C)c1.[Br-]', 'CN(C)C(=O)Oc1ccc[n+](C)c1.[Br-]', 'Cc1ccc2cc3c(ccc4ccccc43)c3c2c1CC3', 'c1ccc2nc3ccccc3cc2c1', 'CNC(=O)Oc1ccc2c(c1)c(C)cn2Nc1ccncc1F', 'COc1ccc2cc3[n+](cc2c1OC)CCc1cc2c(cc1-3)OCO2', 'COc1ccc2cc3[n+](cc2c1OC)CCc1cc2c(cc1-3)OCO2', 'Cc1ccccc1C', 'Cc1ccccc1C', 'Nc1cc(-c2ccccc2)nc2ccccc12', 'O=c1c2ccccc2[se]n1-c1ccccc1', 'C#CCN[C@@H]1CCc2ccccc21', 'O=C1C(=O)c2cc([N+](=O)[O-])ccc2-c2ccccc21', 'O=C1C=Cc2ccccc2C1=O', 'CN(C)C(=O)Oc1cccc([N+](C)(C)C)c1.[Br-]', 'CN(C)C(=O)Oc1cccc([N+](C)(C)C)c1.[Br-]', 'CN(C)C(=O)Oc1cccc([N+](C)(C)C)c1.[Br-]', 'CN(C)C(=O)Oc1cccc([N+](C)(C)C)c1.[Br-]', 'CN(C)C(=O)Oc1cccc([N+](C)(C)C)c1.[Br-]', 'CN(C)C(=O)Oc1cccc([N+](C)(C)C)c1.[Br-]', 'CN(C)C(=O)Oc1cccc([N+](C)(C)C)c1.[Br-]', 'CN(C)C(=O)Oc1cccc([N+](C)(C)C)c1.[Br-]', 'CN(C)C(=O)Oc1cccc([N+](C)(C)C)c1.[Br-]', 'CN(C)C(=O)Oc1cccc([N+](C)(C)C)c1.[Br-]', 'CN(C)C(=O)Oc1cccc([N+](C)(C)C)c1.[Br-]', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C)C12', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C)C12', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C)C12', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C)C12', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C)C12', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C)C12', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C)C12', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2C', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2C', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2C', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)C2N3C)cc1', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)C2N3C)cc1', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)C2N3C)cc1', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)C2N3Cc2ccccc2)cc1', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)C2N3)cc1', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCNC2N3)cc1', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCNC2N3)cc1', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CCc3ccccc3)C1N2C', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCNC1N2C', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2Cc1ccccc1', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCNC2N3C)cc1', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCNC1N2', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(Cc3ccccc3)C1N2Cc1ccccc1', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2', 'O=C(c1ccccc1)N(CCC1CCN(Cc2ccccc2)CC1)c1ccccc1', 'CN(CCC1CCN(Cc2ccccc2)CC1)C(=O)c1ccccc1', 'O=C(NCCC1CCN(Cc2ccccc2)CC1)c1ccccc1', 'C/C=C1\\C2C=C(C(F)(F)F)CC1(N)c1ccc(O)nc1C2', 'CC1=CC2Cc3nc(O)ccc3C(N)(C1)/C2=C/C(F)(F)F', 'Cc1ccc(N)cc1', 'Cc1ccc(N)cc1', 'NC12CC(C(F)(F)F)=CC(Cc3nc(O)ccc31)/C2=C\\C(F)(F)F', 'CC[N+](C)(C)c1cccc(O)c1.[Br-]', 'CC[N+](C)(C)c1cccc(O)c1.[Br-]', 'CC[N+](C)(C)c1cccc(O)c1.[Br-]', 'CC[N+](C)(C)c1cccc(O)c1.[Br-]', 'C=CC[N+](C)(C)c1ccccc1O.[Br-]', 'C=CC[N+](C)(C)c1ccccc1OC.[Br-]', 'C=CC[N+](C)(C)c1cccc(O)c1.[Br-]', 'COc1ccc(C(=O)COc2ccc([N+](C)(C)C)cc2)cc1OC.[Br-]', 'COc1ccc(C(=O)COc2cccc([N+](C)(C)Cc3ccccc3)c2)cc1OC.[Br-]', 'COc1ccc(C(=O)COc2cccc([N+](C)(C)C)c2)cc1OC.[Br-]', 'C=CCC[N+](C)(C)c1ccccc1O.[Br-]', 'COc1ccc(C(=O)CCc2cccc([N+](C)(C)Cc3ccccc3)c2)cc1OC.[Br-]', 'C=CC[N+](C)(C)c1ccc(OCC(=O)c2ccc(OC)c(OC)c2)cc1.[Br-]', 'COc1ccc(C(=O)COc2ccc([N+](C)(C)Cc3ccccc3)cc2)cc1OC.[Br-]', 'COc1ccc(C(=O)CCc2ccc([N+](C)(C)Cc3ccccc3)cc2)cc1OC.[Br-]', 'COc1ccc(C(=O)CCc2ccc([N+](C)(C)C)cc2)cc1OC.[Br-]', 'C=CC[N+](C)(C)c1cccc(OC(C)=O)c1.[Br-]', 'CCCCCCCNC(=O)Oc1ccc2c(c1)C1(C)CCN(CC)C2C1', 'Nc1ccc2nc3c(c(N)c2c1)CCCC3', 'Nc1ccc2nc3c(c(N)c2c1)CCCC3', 'Nc1ccc2nc3c(c(N)c2c1)CCCC3', 'COc1ccc2c(N)c3c(nc2c1)CCCC3', 'COc1ccc2c(N)c3c(nc2c1)CCCC3', 'COc1ccc2c(N)c3c(nc2c1)CCCC3', 'COc1ccc2c(N)c3c(nc2c1)CCCC3', 'COc1ccc2c(c1)Cc1c-2nc2ccccc2c1N', 'CCCN1CCCC(c2cccc(OC(=O)NC(C)C)c2)C1', 'COc1cccc2c1Cc1c-2nc2ccccc2c1N', 'Nc1c2c(nc3cc(F)ccc13)-c1ccccc1C2', 'Nc1c2c(nc3cc(Cl)ccc13)-c1ccccc1C2', 'Nc1ccc2c(N)c3c(nc2c1)-c1ccccc1C3', 'Nc1ccc2nc3c(c(N)c2c1)Cc1ccccc1-3', 'Nc1c2c(nc3ccccc13)-c1ccccc1C2', 'Nc1c2c(nc3ccccc13)-c1ccccc1C2', 'Nc1c2c(nc3ccccc13)-c1ccccc1C2', 'O=[N+]([O-])c1ccc2c(NCc3ccccc3)c3c(nc2c1)CCCC3', 'O=[N+]([O-])c1ccc2c(NCc3ccccc3)c3c(nc2c1)CCCC3', 'CCCCCCCNc1c2c(nc3cc(F)ccc13)CCCC2', 'CCCCCCCNc1c2c(nc3cc(F)ccc13)CCCC2', 'CCCCCCCNc1c2c(nc3cc(F)ccc13)CCCC2', 'CCCCCCCNc1c2c(nc3cc(OC)ccc13)CCCC2', 'CCCCCCCNc1c2c(nc3cc(OC)ccc13)CCCC2', 'CCCCCCCNc1c2c(nc3cc(OC)ccc13)CCCC2', 'CCCCCCCNc1c2c(nc3cc([N+](=O)[O-])ccc13)CCCC2', 'CCCCCCCNc1c2c(nc3cc([N+](=O)[O-])ccc13)CCCC2', 'CCCCCCCNc1c2c(nc3cc([N+](=O)[O-])ccc13)CCCC2', 'CCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'CCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'CCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'CCCCCCNC(=O)Oc1ccc2c(c1)C1CCN(CC)C2C1', 'CCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CC)[C@@H]1C2', 'CCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CC)[C@@H]1C2', 'CCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CC)[C@@H]1C2', 'CCCCCCCNc1c2c(nc3cc(C)ccc13)CCCC2', 'CCCCCCCNc1c2c(nc3cc(C)ccc13)CCCC2', 'CCCCCCCNc1c2c(nc3cc(C)ccc13)CCCC2', 'CCN1CCC2CC1c1ccc(OC(=O)NC)cc12', 'CCCCNC(=O)Oc1ccc2c(c1)C1CCN(CC)C2C1', 'CCCCCCNC(=O)Oc1ccc2c(c1)C1CCN(C)C2C1', 'CCCCCCCNC(=O)Oc1ccc2c(c1)C1CCN(CC)C2C1', 'Cc1ccc2c(NCc3ccccc3)c3c(nc2c1)CCCC3', 'Cc1ccc2c(NCc3ccccc3)c3c(nc2c1)CCCC3', 'CCCCCCCNC(=O)Oc1ccc2c(c1)C1CCN(C)C2C1', 'CCCCCCNC(=O)Oc1ccc2c(c1)C1(C)CCN(CC)C2C1', 'Nc1c2c(nc3cc([N+](=O)[O-])ccc13)CCCC2', 'Nc1c2c(nc3cc([N+](=O)[O-])ccc13)CCCC2', 'Nc1c2c(nc3cc([N+](=O)[O-])ccc13)CCCC2', 'Nc1c2c(nc3cc([N+](=O)[O-])ccc13)CCCC2', 'O=[N+]([O-])c1ccc2nc3c(c(NCc4ccccc4)c2c1)CCCC3', 'O=[N+]([O-])c1ccc2nc3c(c(NCc4ccccc4)c2c1)CCCC3', 'Clc1ccc2c(NCc3ccccc3)c3c(nc2c1)CCCC3', 'Clc1ccc2c(NCc3ccccc3)c3c(nc2c1)CCCC3', 'Cc1ccc2nc3c(c(NCc4ccccc4)c2c1)CCCC3', 'Cc1ccc2nc3c(c(NCc4ccccc4)c2c1)CCCC3', 'Nc1c2c(nc3ccc(Cl)cc13)CCCC2', 'Nc1c2c(nc3ccc(Cl)cc13)CCCC2', 'Nc1c2c(nc3ccc(Cl)cc13)CCCC2', 'Nc1c2c(nc3ccc(Cl)cc13)CCCC2', 'Nc1c2c(nc3ccc(Cl)cc13)CCCC2', 'Nc1c2c(nc3ccc(Cl)cc13)CCCC2', 'Nc1c2c(nc3ccc(Cl)cc13)CCCC2', 'Cc1ccc2nc3c(c(N)c2c1)CCCC3', 'Cc1ccc2nc3c(c(N)c2c1)CCCC3', 'Cc1ccc2nc3c(c(N)c2c1)CCCC3', 'Cc1ccc2nc3c(c(N)c2c1)CCCC3', 'Nc1c2c(nc3ccccc13)-c1ccc(F)cc1C2', 'Cc1cccc2c1Cc1c-2nc2ccccc2c1N', 'COc1cc2nc3c(c(N)c2cc1OC)CCCC3', 'COc1cc2nc3c(c(N)c2cc1OC)CCCC3', 'COc1cc2nc3c(c(N)c2cc1OC)CCCC3', 'COc1cc2nc3c(c(N)c2cc1OC)CCCC3', 'Nc1c2c(nc3cc(Cl)c(Cl)cc13)CCCC2', 'Nc1c2c(nc3cc(Cl)c(Cl)cc13)CCCC2', 'Nc1c2c(nc3cc(Cl)c(Cl)cc13)CCCC2', 'Nc1c2c(nc3cc(Cl)c(Cl)cc13)CCCC2', 'Nc1c2c(nc3cc(F)ccc13)CCCC2', 'Nc1c2c(nc3cc(F)ccc13)CCCC2', 'Nc1c2c(nc3cc(F)ccc13)CCCC2', 'Nc1c2c(nc3cc(F)ccc13)CCCC2', 'Nc1c2c(nc3cc([N+](=O)[O-])ccc13)-c1ccccc1C2', 'COc1ccc2c(c1)-c1nc3ccccc3c(N)c1C2', 'CCCCCCCNc1c2c(nc3ccc(C)cc13)CCCC2', 'CCCCCCCNc1c2c(nc3ccc(C)cc13)CCCC2', 'CCCCCCCNc1c2c(nc3ccc(C)cc13)CCCC2', 'Nc1c2c(nc3cc(Br)ccc13)CCCC2', 'Nc1c2c(nc3ccccc13)-c1ccc(Cl)cc1C2', 'Cc1ccc2c(N)c3c(nc2c1)CCCC3', 'Cc1ccc2c(N)c3c(nc2c1)CCCC3', 'Cc1ccc2c(N)c3c(nc2c1)CCCC3', 'Cc1ccc2c(N)c3c(nc2c1)CCCC3', 'Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3ccc([N+](=O)[O-])cc13)CCCC2', 'Nc1c2c(nc3ccc([N+](=O)[O-])cc13)CCCC2', 'Nc1c2c(nc3ccc([N+](=O)[O-])cc13)CCCC2', 'Nc1c2c(nc3ccc([N+](=O)[O-])cc13)CCCC2', 'Cc1ccc2c(c1)-c1nc3ccccc3c(N)c1C2', 'COc1ccccc1CNCCCCCCNCCCCCCNCCCCCCNCc1ccccc1OC', 'CN(C)c1ccc(C(=C2C=CC(=[N+](C)C)C=C2)c2ccc(N(C)C)cc2)cc1.[Cl-]', 'CN1CC[C@@]2(C)c3cc(OC(=O)NCCN4CCOCC4)ccc3N(C)[C@@H]12', 'CN1CC[C@@]2(C)c3cc(OC(=O)NCCCCCCCN4CCOCC4)ccc3N(C)[C@@H]12', 'CN1CC[C@@]2(C)c3cc(OC(=O)NCCCCCCCCCN4CCOCC4)ccc3N(C)[C@@H]12', 'CN1CC[C@@]2(C)c3cc(OC(=O)NCCCCCCN4CCOCC4)ccc3N(C)[C@@H]12', 'CN1CC[C@@]2(C)c3cc(OC(=O)NCCCN4CCOCC4)ccc3N(C)[C@@H]12', 'CN1CC[C@@]2(C)c3cc(OC(=O)NCCCCCCCCCCCN4CCOCC4)ccc3N(C)[C@@H]12', 'CN1CC[C@@]2(C)c3cc(OC(=O)NCCCCCCCCCCN4CCOCC4)ccc3N(C)[C@@H]12', 'CN1CC[C@@]2(C)c3cc(OC(=O)NCCCCCCCCN4CCOCC4)ccc3N(C)[C@@H]12', 'CN1CC[C@@]2(C)c3cc(OC(=O)NCCCCCN4CCOCC4)ccc3N(C)[C@@H]12', 'CN1CC[C@@]2(C)c3cc(OC(=O)NCCCCN4CCOCC4)ccc3N(C)[C@@H]12', 'CN1CC[C@@]2(C)c3cc(OC(=O)NCCCCCCCCCCCCN4CCOCC4)ccc3N(C)[C@@H]12', 'O=c1c(O)c(-c2cc(O)c(O)c(O)c2)oc2cc(O)cc(O)c12', 'O=c1c(O)c(-c2cc(O)c(O)c(O)c2)oc2cc(O)cc(O)c12', 'CN(Cc1ccc(C(C)(C)C)cc1)Cc1cccc2ccccc12', 'CCCCCCCCc1noc(/C(=N/O)SCCN(CC)CC)n1', 'CCN(CC)CCS/C(=N\\O)c1nc(Cc2ccccc2)no1', 'COc1cc2ccc(=O)oc2cc1O', 'CC(C)(C)c1ccc(C(=O)CCCN2CCC(OC(c3ccccc3)c3ccccc3)CC2)cc1', 'Fc1ccc2c(NCCCCCCCNc3c4c(nc5cc(F)ccc35)CCCC4)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCCCCCCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'c1cnc2nc3c(c(NCCCCCCCCNc4c5c(nc6ncccc46)CCCC5)c2c1)CCCC3', 'COc1ccc(CN2CCC(CCC(=O)c3ccc4c(c3)CCN4)CC2)cc1', 'COc1ccccc1CN1CCC(CCC(=O)c2ccc3c(c2)CCN3)CC1', 'CCCCCN1CCc2cc(C(=O)CCC3CCN(Cc4ccccc4)CC3)ccc21', 'CN1CCc2cc(C(=O)CCC3CCN(Cc4ccccc4)CC3)ccc21', 'c1ccc2c(NCCCCCCCCNc3c4c(nc5ccccc35)CCC4)c3c(nc2c1)CCC3', 'c1ccc2c(NCCCCCCCNc3c4c(nc5ccccc35)CCC4)c3c(nc2c1)CCC3', 'c1ccc2c(NCCCCCCCCNc3c4c(nc5ccccc35)CCCCC4)c3c(nc2c1)CCCCC3', 'c1cnc2nc3c(c(NCCCCCCCNc4c5c(nc6ncccc46)CCCC5)c2c1)CCCC3', 'c1ccc2c(NCCCCCCCNc3c4c(nc5ccccc35)CCCCC4)c3c(nc2c1)CCCCC3', 'c1cnc2nc3c(c(NCCCCCCNc4c5c(nc6ncccc46)CCCC5)c2c1)CCCC3', 'Clc1ccc2c(NCCCCCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCCCCCNc3c4c(nc5ccccc35)CCC4)c3c(nc2c1)CCC3', 'Clc1ccc2c(NCCCCCCCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)c3c(nc2c1)CCCC3', 'Fc1ccc2c(NCCCCCCNc3c4c(nc5cc(F)ccc35)CCCC4)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCCCCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'Clc1ccc2c(NCCCCCCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)c3c(nc2c1)CCCC3', 'O=C(CCC1CCN(Cc2cccc(F)c2)CC1)c1ccc2c(c1)CCN2', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)CCN2', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)CCN2', 'Fc1ccc2c(NCCCCCCCCNc3c4c(nc5cc(F)ccc35)CCCC4)c3c(nc2c1)CCCC3', 'COc1cc2cc(-c3ccc(CN(C)Cc4ccccc4)cc3)c(=O)oc2cc1OC', 'COc1cc2cc(-c3ccc(CN(C)Cc4ccccc4)cc3)c(=O)oc2cc1OC', 'COc1cc2cc(-c3ccc(CN(C)Cc4ccccc4)cc3)c(=O)oc2cc1OC', 'COc1cc2cc(-c3ccc(CN(C)Cc4ccccc4)cc3)c(=O)oc2cc1OC', 'COc1cc2cc(-c3ccc(CN(C)Cc4ccccc4)cc3)c(=O)oc2cc1OC', 'COc1cc2cc(-c3ccc(CN(C)Cc4ccccc4)cc3)c(=O)oc2cc1OC', 'COc1cc2cc(-c3ccc(CN(C)Cc4ccccc4)cc3)c(=O)oc2cc1OC', 'COc1cc2cc(-c3ccc(CN(C)Cc4ccccc4)cc3)c(=O)oc2cc1OC', 'COc1cc2cc(-c3ccc(CN(C)Cc4ccccc4)cc3)c(=O)oc2cc1OC', 'COc1cc2cc(-c3ccc(CN(C)Cc4ccccc4)cc3)c(=O)oc2cc1OC', 'O=C(CCC1CCN(Cc2ccccc2F)CC1)c1ccc2c(c1)CCN2', 'O=C(CCC1CCN(Cc2ccccc2Cl)CC1)c1ccc2c(c1)CCN2', 'O=C(CCC1CCN(Cc2ccc(Cl)cc2)CC1)c1ccc2c(c1)CCN2', 'CN1CCCc2ccc(C(=O)CCC3CCN(Cc4ccccc4)CC3)cc21', 'O=C(CCC1CCN(Cc2cccc(Cl)c2)CC1)c1ccc2c(c1)CCN2', 'O=C(CCC1CCN(Cc2ccc(F)cc2)CC1)c1ccc2c(c1)CCN2', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)N(Cc1ccccc1)CC2', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)NCC2', 'O=C(CCC1CCN(Cc2cccc([N+](=O)[O-])c2)CC1)c1ccc2c(c1)CCN2', 'Cc1cccc(CN2CCC(CCC(=O)c3ccc4c(c3)CCN4)CC2)c1', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)NCCCCC2', 'O=C(CCC1CCN(Cc2cccc([N+](=O)[O-])c2)CC1)c1ccc2c(c1)NCCCC2', 'Cc1cccc(CN2CCC(CCC(=O)c3ccc4c(c3)NCCCC4)CC2)c1', 'O=C(CCC1CCN(Cc2cccc(Cl)c2)CC1)c1ccc2c(c1)NCCCC2', 'O=C(CCC1CCN(Cc2ccc(F)cc2)CC1)c1ccc2c(c1)NCCCC2', 'O=C(CCC1CCN(Cc2cccc(F)c2)CC1)c1ccc2c(c1)NCCCC2', 'CCN1CCc2cc(C(=O)CCC3CCN(Cc4ccccc4)CC3)ccc21', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)CCCCN2', 'CCCN1CCc2cc(C(=O)CCC3CCN(Cc4ccccc4)CC3)ccc21', 'CCCCN1CCc2cc(C(=O)CCC3CCN(Cc4ccccc4)CC3)ccc21', 'COc1cccc(CN2CCC(CCC(=O)c3ccc4c(c3)CCN4)CC2)c1', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)NCCC2', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)NCCC2', 'O=C(CCC1CCN(Cc2ccccc2F)CC1)c1ccc2c(c1)NCCCC2', 'COc1ccc(CN2CCC(CCC(=O)c3ccc4c(c3)NCCCC4)CC2)cc1', 'COc1cccc(CN2CCC(CCC(=O)c3ccc4c(c3)NCCCC4)CC2)c1', 'COc1ccccc1CN1CCC(CCC(=O)c2ccc3c(c2)NCCCC3)CC1', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)N(Cc1ccccc1)CCCC2', 'CCCN1CCCCc2ccc(C(=O)CCC3CCN(Cc4ccccc4)CC3)cc21', 'CCN1CCCCc2ccc(C(=O)CCC3CCN(Cc4ccccc4)CC3)cc21', 'CN1CCCCc2ccc(C(=O)CCC3CCN(Cc4ccccc4)CC3)cc21', 'O=c1c(O)c(-c2ccccc2)oc2cc(O)cc(O)c12', 'O=c1c(O)c(-c2ccccc2)oc2cc(O)cc(O)c12', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCNC1N2C', 'CN1CC[C@@]2(C)c3cc(OC(=O)NC(C)(C)C)ccc3N(C)[C@@H]12', 'CN1CC[C@@]2(C)c3cc(OC(=O)NC(C)(C)C)ccc3N(C)[C@@H]12', 'CC(C)OP(=O)(F)OC(C)C', 'CC(C)OP(=O)(F)OC(C)C', 'CC(C)OP(=O)(F)OC(C)C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(Cc3ccccc3)C1N2C', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C)[C@@H]12', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C)[C@@H]12', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C)[C@@H]12', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C)[C@@H]12', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C)[C@@H]12', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C)[C@@H]12', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C)[C@@H]12', 'CN(C)C(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CN(C)C(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CN(C)C(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CCCCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CCCCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(Cc3ccccc3)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(Cc3ccccc3)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(Cc3ccccc3)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN[C@@H]1N2C', 'C=CCN1CC[C@@]2(C)c3cc(OC(=O)NC)ccc3N(C)[C@@H]12', 'C=CCN1CC[C@@]2(C)c3cc(OC(=O)NC)ccc3N(C)[C@@H]12', 'COc1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)[C@@H]2N3C)cc1', 'COc1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)[C@@H]2N3C)cc1', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CCc3ccccc3)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CCc3ccccc3)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@@]1(C)CCNC1N2', 'C[C@]12CCNC1Nc1ccc(OC(=O)Nc3ccccc3)cc12', 'CNC(=O)Oc1ccc2c(c1)[C@@]1(C)CCN(C)C1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@@]1(C)CCN(C)C1N2C', 'CN1CC[C@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C)C12', 'CN1CC[C@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C)C12', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(Cc3ccccc3)C1N2Cc1ccccc1', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3NC12', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3NC12', 'CNC(=O)Oc1ccc2c(c1)[C@@]1(C)CCN(Cc3ccccc3)C1N2Cc1ccccc1', 'CN1c2ccc(OC(=O)Nc3ccccc3)cc2[C@]2(C)CCNC12', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2Cc1ccccc1', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2Cc1ccccc1', 'C[C@@]12CCN(Cc3ccccc3)C1N(Cc1ccccc1)c1ccc(OC(=O)Nc3ccccc3)cc12', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCNC1N2', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2', 'C[C@@]12CCNC1Nc1ccc(OC(=O)Nc3ccccc3)cc12', 'CCCCN1CCC(CCC(=O)c2cc(Cl)c(N)cc2OC)CC1', 'CCCCN1CCC(CCC(=O)c2cc(Cl)c(N)cc2OC)CC1', 'COCC[N+]1(C)CCC(CCC(=O)c2cc3cc(OC)c(OC)cc3s2)CC1.[Br-]', 'COc1cc2cc(CC3CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(O)C3CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)CC3=CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)CC3(O)CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)CC3(O)CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)CCC3CC[N+](C)(Cc4ccc([N+](=O)[O-])o4)CC3)sc2cc1OC.[Br-]', 'COC(=O)C[N+]1(C)CCC(CCC(=O)c2cc3cc(OC)c(OC)cc3s2)CC1.[Br-]', 'COc1cc2cc(C=C3CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'COc1cc2c(cc1OC)C(=O)C(Cc1cc[n+](Cc3ccccc3)cc1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CC[N+](C)(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CC[N+](C)(Cc3ccccc3)CC1)C2', 'C[N+]1(Cc2ccccc2)CCC(CCC(=O)c2ccc3c(c2)NCCCC3)CC1', 'COc1cc2c(cc1OC)C(=O)[C@@H]([C@@H](O)C1CC[N+](C)(Cc3ccccc3)CC1)C2', 'Cc1cc(C(=O)CCC2CC[N+](C)(Cc3ccccc3)CC2)nn1-c1ccccc1', 'Cc1cc(C(=O)CCC2CC[N+](C)(Cc3ccccc3)CC2)nn1-c1ccccc1', 'Cc1cc(C(=O)CCC2CC[N+](C)(Cc3ccccc3)CC2)nn1-c1ccccc1', 'Cc1cc(C(=O)CCC2CC[N+](C)(Cc3ccccc3)CC2)nn1-c1ccccc1', 'COc1cc2cc(C(=O)C=C3CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)C=C3CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)CCc3cc[n+](CC4CCC4)cc3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)CCc3cc[n+](CC4CCC4)cc3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)CC(O)C3CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)CC(O)C3CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)CC(O)C3CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)CC(O)C3CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)CC(O)C3CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)CC(OC)C3CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)CC(OC)C3CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'C=CC[n+]1ccc(CCC(=O)c2cc3cc(OC)c(OC)cc3s2)cc1.[Br-]', 'COc1cc2cc(C(=O)CCc3cc[n+](CC4CC4)cc3)sc2cc1OC.[Br-]', 'CCC[n+]1ccc(CCC(=O)c2cc3cc(OC)c(OC)cc3s2)cc1.[Br-]', 'CC[n+]1ccc(CCC(=O)c2cc3cc(OC)c(OC)cc3s2)cc1.[Br-]', 'COc1cc2cc(C(=O)CCC3CC[N+](C)(Cc4ccccc4)CC3)n(C)c2cc1OC', 'COc1cc2cc(C(=O)CCC3CC[N+](C)(Cc4ccccc4)CC3)n(C)c2cc1OC', 'COc1cc2cc(C(=O)CCC3CC[N+](C)(Cc4ccccc4)CC3)n(C)c2cc1OC', 'COc1cc2cc(C(=O)CCC3CC[N+](C)(Cc4ccccc4)CC3)n(C)c2cc1OC', 'COc1cc2cc(C(=O)CCc3cc[n+](C)cc3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)CCc3cc[n+](C)cc3)sc2cc1OC.[Br-]', 'COc1cc2c(cc1OC)C(=O)C(C(O)C1CC[N+](C)(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)[C@@H]([C@H](O)C1CC[N+](C)(Cc3ccccc3)CC1)C2', 'C[N+]1(Cc2ccccc2)CCC(CCC(=O)c2ccc3c(c2)OCCO3)CC1', 'C[N+]1(Cc2ccccc2)CCC(CCC(=O)c2ccc3c(c2)OCCO3)CC1', 'C[N+]1(Cc2ccccc2)CCC(CCC(=O)c2ccc3c(c2)OCCO3)CC1', 'COc1cc2cc(C(=O)CCCCC3CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)CC(O)CCC3CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)CC(O)CCC3CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1cc[n+](Cc3ccccc3)cc1)C2', 'COc1ccc2c3c1OC1C[C@@H](O)C=C[C@@]31CC[N+](C)(C)C2', 'COc1cc2c(cc1OC)C(=O)C(C(O)c1cc[n+](Cc3ccccc3)cc1)C2', 'COc1cc2cc(C(=O)CCC3CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)CCC3CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC.[Br-]', 'COc1cc2c(cc1OC)C(=O)C(C(O)C1CC[N+](Cc3ccccc3)(Cc3ccccc3)CC1)C2', 'COc1cc2cc(C(=O)CCc3cc[n+](Cc4ccsc4)cc3)sc2cc1OC', 'COc1cc2cc(C(=O)CCc3cc[n+](Cc4ccsc4)cc3)sc2cc1OC', 'COC(=O)c1ccc(C[n+]2ccc(CCC(=O)c3cc4cc(OC)c(OC)cc4s3)cc2)cc1', 'COC(=O)c1ccc(C[n+]2ccc(CCC(=O)c3cc4cc(OC)c(OC)cc4s3)cc2)cc1', 'COc1cc2cc(C(=O)CCc3cc[n+](Cc4ccc(F)cc4)cc3)sc2cc1OC', 'COc1cc2cc(C(=O)CCc3cc[n+](Cc4ccc(F)cc4)cc3)sc2cc1OC', 'COc1cc2cc(C(=O)CCC3CC[N+](C)(CC(=O)OC(C)(C)C)CC3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)CCC3CC[N+](C)(CCOc4ccccc4)CC3)sc2cc1OC.[Br-]', 'C[N+]1(Cc2ccccc2)CCC(CCC(=O)c2ccc3c(c2)OCO3)CC1', 'C[N+]1(Cc2ccccc2)CCC(CCC(=O)c2ccc3c(c2)OCO3)CC1', 'C[N+]1(Cc2ccccc2)CCC(CCC(=O)c2ccc3c(c2)OCO3)CC1', 'C[N+]1(Cc2ccccc2)CCC(CCC(=O)c2ccc3c(c2)OCO3)CC1', 'COc1ccc2cc(C(=O)CCC3CC[N+](C)(Cc4ccccc4)CC3)ccc2c1', 'COc1ccc2cc(C(=O)CCC3CC[N+](C)(Cc4ccccc4)CC3)ccc2c1', 'COc1ccc2cc(C(=O)CCC3CC[N+](C)(Cc4ccccc4)CC3)ccc2c1', 'COc1ccc2cc(C(=O)CCC3CC[N+](C)(Cc4ccccc4)CC3)ccc2c1', 'COc1cc2cc(C(=O)CCC3CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC', 'COc1cc2cc(C(=O)CCC3CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC', 'C[N+]1(Cc2ccccc2)CCC(CCC(=O)c2ccc3c(c2)CCCCO3)CC1', 'C[N+]1(Cc2ccccc2)CCC(CCC(=O)c2ccc3c(c2)CCCCO3)CC1', 'C[N+]1(Cc2ccccc2)CCC(CCC(=O)c2ccc3c(c2)CCCCO3)CC1', 'COc1cc2cc(C(=O)CCc3cc[n+](Cc4ccccc4)cc3)sc2cc1OC', 'COc1cc2cc(C(=O)CCc3cc[n+](Cc4ccccc4)cc3)sc2cc1OC', 'C[N@+]1(Cc2ccccc2)CC[C@H](CCC(=O)c2ccc3c(c2)CCO3)CC1', 'C[N+]1(Cc2ccccc2)CCC(CCC(=O)c2ccc3c(c2)CCO3)CC1', 'C[N+]1(Cc2ccccc2)CCC(CCC(=O)c2ccc3c(c2)CCO3)CC1', 'CCOCC[n+]1ccc(CCC(=O)c2cc3cc(OC)c(OC)cc3s2)cc1', 'COCC[n+]1ccc(CCC(=O)c2cc3cc(OC)c(OC)cc3s2)cc1', 'COCC[n+]1ccc(CCC(=O)c2cc3cc(OC)c(OC)cc3s2)cc1', 'COc1cc2cc(C(=O)CCC3CC[N+](C)(CC4CCCCO4)CC3)sc2cc1OC.[Br-]', 'Nc1c2c(nc3ccccc13)-c1ccccc1CC2', 'Nc1c2c(nc3ccccc13)-c1ccccc1CC2', 'Nc1c2c(nc3ccccc13)-c1ccccc1CS2', 'CC(C)(C)c1cccc(C(=O)C(F)(F)F)c1', 'CC(C)(C)c1cccc(C(=O)C(F)(F)F)c1', 'CCCc1cc(=NCCC2CCN(Cc3ccccc3)CC2)[nH]nc1-c1ccccc1', 'CCc1cc(=NCCC2CCN(Cc3ccccc3)CC2)[nH]nc1-c1ccccc1', 'c1ccc(CN2CCC(CCN=c3[nH]nc(-c4ccccc4)c4ccccc34)CC2)cc1', 'CC(C)c1cc(-c2ccccc2)n[nH]c1=NCCC1CCN(Cc2ccccc2)CC1', 'Cc1cc(=NCCC2CCN(Cc3ccccc3)CC2)[nH]nc1-c1ccccc1', 'Cc1cc(-c2ccccc2)n[nH]c1=NCCC1CCN(Cc2ccccc2)CC1', 'Cc1ccc2onc(CCC3CCN(Cc4ccccc4)CC3)c2c1', 'Cc1ccc2onc(CCC3CCN(Cc4ccccc4)CC3)c2c1', 'Cc1ccc2onc(CCC3CCN(Cc4ccccc4)CC3)c2c1', 'Cc1ccc2onc(CCC3CCN(Cc4ccccc4)CC3)c2c1', 'Cc1ccc2onc(CCC3CCN(Cc4ccccc4)CC3)c2c1', 'Cc1ccc2onc(CCC3CCN(Cc4ccccc4)CC3)c2c1', 'Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'Oc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'Oc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'Oc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'Oc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'Oc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'Oc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'N#Cc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'N#Cc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'N#Cc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'N#Cc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'N#Cc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'Brc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'Brc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'Brc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'Brc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'Brc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'Brc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'c1ccc(CN2CCC(CCc3n[nH]c4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3n[nH]c4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3n[nH]c4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3n[nH]c4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3n[nH]c4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3ncnc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3ncnc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3ncnc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3ncnc4ccccc34)CC2)cc1', 'O=S(=O)(Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1)c1ccccc1', 'O=S(=O)(Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1)c1ccccc1', 'O=S(=O)(Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1)c1ccccc1', 'O=S(=O)(Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1)c1ccccc1', 'O=S(=O)(Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1)c1ccccc1', 'O=C(Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1)c1ccccc1', 'O=C(Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1)c1ccccc1', 'O=C(Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1)c1ccccc1', 'O=C(Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1)c1ccccc1', 'O=C(Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1)c1ccccc1', 'O=C(Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1)c1ccccc1', 'C(=C\\C1CCN(Cc2ccccc2)CC1)\\c1noc2ccccc12', 'c1ccc(CN2CCC(CCc3nccc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3nccc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3nccc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3nccc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3nccc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3nccc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3noc4cc(N5CCOCC5)ccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3noc4cc(N5CCOCC5)ccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3noc4cc(N5CCOCC5)ccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3noc4cc(N5CCOCC5)ccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3noc4cc(N5CCOCC5)ccc34)CC2)cc1', 'COc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'COc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'COc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'COc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'COc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'COc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'CC(=O)Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'CC(=O)Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'CC(=O)Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'CC(=O)Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'CC(=O)Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'CC(=O)Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'c1ccc(CN2CCC(CCc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3noc4ccccc34)CC2)cc1', 'COc1cccc2c(CCC3CCN(Cc4ccccc4)CC3)noc12', 'COc1cccc2c(CCC3CCN(Cc4ccccc4)CC3)noc12', 'COc1cccc2c(CCC3CCN(Cc4ccccc4)CC3)noc12', 'COc1cccc2c(CCC3CCN(Cc4ccccc4)CC3)noc12', 'COc1cccc2c(CCC3CCN(Cc4ccccc4)CC3)noc12', 'COc1cccc2c(CCC3CCN(Cc4ccccc4)CC3)noc12', 'COc1ccc2onc(CCC3CCN(Cc4ccccc4)CC3)c2c1', 'COc1ccc2onc(CCC3CCN(Cc4ccccc4)CC3)c2c1', 'COc1ccc2onc(CCC3CCN(Cc4ccccc4)CC3)c2c1', 'COc1ccc2onc(CCC3CCN(Cc4ccccc4)CC3)c2c1', 'COc1ccc2onc(CCC3CCN(Cc4ccccc4)CC3)c2c1', 'COc1ccc2onc(CCC3CCN(Cc4ccccc4)CC3)c2c1', 'Cc1cc2onc(CCC3CCN(Cc4ccccc4)CC3)c2cc1C', 'Cc1cc2onc(CCC3CCN(Cc4ccccc4)CC3)c2cc1C', 'Cc1cc2onc(CCC3CCN(Cc4ccccc4)CC3)c2cc1C', 'Cc1cc2onc(CCC3CCN(Cc4ccccc4)CC3)c2cc1C', 'c1ccc(CN2CCC(CCc3nsc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3nsc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3nsc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3nsc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCc3nsc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCNc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCNc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCNc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCNc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCNc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CNc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CNc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CNc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CNc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CNc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CNc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(COc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(COc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(COc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(COc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(COc3noc4ccccc34)CC2)cc1', 'NC(=O)c1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'NC(=O)c1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'NC(=O)c1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'NC(=O)c1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'NC(=O)c1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'NC(=O)c1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'NC(=O)c1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'c1ccc(CN2CCC(CCCc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCCc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCCc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCCc3noc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CCCc3noc4ccccc34)CC2)cc1', 'O=C(OC1Cc2c(O)cc(O)cc2OC1c1ccc(O)c(O)c1)c1cc(O)c(O)c(O)c1', 'O=C1Cc2cc3onc(CCC4CCN(Cc5ccccc5)CC4)c3cc2N1', 'O=C1Cc2cc3onc(CCC4CCN(Cc5ccccc5)CC4)c3cc2N1', 'O=C1Cc2cc3onc(CCC4CCN(Cc5ccccc5)CC4)c3cc2N1', 'O=C1Cc2cc3onc(CCC4CCN(Cc5ccccc5)CC4)c3cc2N1', 'O=C1Cc2cc3onc(CCC4CCN(Cc5ccccc5)CC4)c3cc2N1', 'CN1C(=O)Cc2cc3c(CCC4CCN(Cc5ccccc5)CC4)noc3cc21.CS(=O)(=O)O', 'O=C1Cc2c(ccc3c(CCC4CCN(Cc5ccccc5)CC4)noc23)N1', 'O=C1Cc2c(ccc3c(CCC4CCN(Cc5ccccc5)CC4)noc23)N1', 'O=C1Cc2c(ccc3c(CCC4CCN(Cc5ccccc5)CC4)noc23)N1', 'O=C1Cc2c(ccc3c(CCC4CCN(Cc5ccccc5)CC4)noc23)N1', 'O=C1Cc2c(ccc3c(CCC4CCN(Cc5ccccc5)CC4)noc23)N1', 'CC(=O)Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1.O=C(O)/C=C/C(=O)O', 'O=C1CCc2cc3c(CCC4CCN(Cc5ccccc5)CC4)noc3cc2N1', 'O=C1CCc2cc3c(CCC4CCN(Cc5ccccc5)CC4)noc3cc2N1', 'O=C1CCc2cc3c(CCC4CCN(Cc5ccccc5)CC4)noc3cc2N1', 'O=C1CCc2cc3c(CCC4CCN(Cc5ccccc5)CC4)noc3cc2N1', 'O=C1CCc2cc3c(CCC4CCN(Cc5ccccc5)CC4)noc3cc2N1', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2C=O', 'CN1CC[C@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3NC12', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(Cc3ccccc3)C12', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C=O)C12', 'CN1CC[C@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(Cc3ccccc3)C12', 'CNC(=O)Oc1ccc2c(c1)[C@@]1(C)CCN(C)C1N2Cc1ccccc1', 'CNC(=O)Oc1ccc2c(c1)[C@@]1(C)CCN(C)C1N2', 'CN(CCOCCNC(=O)Nc1ccc([N+](=O)[O-])cc1)Cc1ccccc1', 'CN(CCOCCNC(=S)NC(=O)c1ccc(-c2ccccc2)cc1)Cc1ccccc1', 'CN(CCOCCNC(=S)NC(=O)c1ccc(Cl)c(Cl)c1)Cc1ccccc1', 'CN(CCOCCNC(=S)Nc1ccc(OC(F)(F)F)cc1)Cc1ccccc1', 'CN(CCOCCNC(=S)Nc1cccc2ccccc12)Cc1ccccc1', 'CN(CCOCCNC(=S)NC1CCCCC1)Cc1ccccc1', 'CN(CCOCCNC(=S)Nc1ccc(Cl)cc1)Cc1ccccc1', 'COc1ccc(NC(=S)NCCOCCN(C)Cc2cccc([N+](=O)[O-])c2)cc1OC', 'CN(CCOCCNC(=S)NC(=O)c1cccc([N+](=O)[O-])c1)Cc1ccccc1', 'CCN(CCOCCNC(=O)NCc1ccccc1)CC1CCCCC1', 'CN(CCOCCNC(=S)NC(=O)c1ccc(Cl)cc1)Cc1ccccc1', 'COc1ccc(C(=O)NC(=S)NCCOCCN(C)Cc2ccccc2)cc1', 'Cc1ccc(C(=O)NC(=S)NCCOCCN(C)Cc2ccccc2)cc1', 'CN(CCOCCNC(=S)NC(=O)c1ccc2ccccc2c1)Cc1ccccc1', 'CN(CCOCCNC(=S)NC(=O)c1ccccc1)Cc1ccccc1', 'CN(CCOCCNC(=S)NC(=O)c1ccccc1)C1CCCCC1', 'CN(CCOCCCNC(=S)NC(=O)c1ccccc1)Cc1ccccc1', 'CC(C)CCSC(=O)OCC[N+](C)(C)C.[Cl-]', 'CC(C)CCSC(=O)OCC[N+](C)(C)C.[Cl-]', 'CCCCCCSC(=O)OCC[N+](C)(C)C.[Cl-]', 'CCCCCCSC(=O)OCC[N+](C)(C)C.[Cl-]', 'CCC(C)(C)SC(=O)OCC[N+](C)(C)C.[Cl-]', 'Nc1c2c(nc3ccccc13)CCC(O)C2', 'Nc1c2c(nc3ccccc13)CCC(O)C2', 'Nc1c2c(nc3ccccc13)CC(O)CC2', 'Nc1c2c(nc3ccccc13)CC(O)CC2', 'Nc1c2c(nc3ccccc13)C(O)CCC2', 'Nc1c2c(nc3ccccc13)C(O)CCC2', 'CCCCCSC(=O)OCC[N+](C)(C)C.[Cl-]', 'CCCCCSC(=O)OCC[N+](C)(C)C.[Cl-]', 'CCCCCSC(=O)OCC[N+](C)(C)C.[Cl-]', 'CCCCCSC(=O)OCC[N+](C)(C)C.[Cl-]', 'CCCCCCCSC(=O)OCC[N+](C)(C)C.[Cl-]', 'CCCCCCCSC(=O)OCC[N+](C)(C)C.[Cl-]', 'CCCCCCCSC(=O)OCC[N+](C)(C)C.[Cl-]', 'CC[N+](C)(C)c1cccc(O)c1.[Cl-]', 'CC[N+](C)(C)c1cccc(O)c1.[Cl-]', 'CC[N+](C)(C)c1cccc(O)c1.[Cl-]', 'CC[N+](C)(C)c1cccc(O)c1.[Cl-]', 'CC[N+](C)(C)c1cccc(O)c1.[Cl-]', 'CC[N+](C)(C)c1cccc(O)c1.[Cl-]', 'CC[N+](C)(C)c1cccc(O)c1.[Cl-]', 'CC[N+](C)(C)c1cccc(O)c1.[Cl-]', 'CC[N+](C)(C)c1cccc(O)c1.[Cl-]', 'CC[N+](C)(C)c1cccc(O)c1.[Cl-]', 'CC[N+](C)(C)c1cccc(O)c1.[Cl-]', 'CC[N+](C)(C)c1cccc(O)c1.[Cl-]', 'CCc1ccccc1SC(=O)OCC[N+](C)(C)C.[Cl-]', 'CCc1ccc(SC(=O)OCC[N+](C)(C)C)cc1.[Cl-]', '[Br-].[Br-].c1ccc2c[n+](CCCCCCCCCCC[n+]3ccc4ccccc4c3)ccc2c1', '[Br-].[Br-].c1cc[n+](CCCCCCCCC[n+]2ccccc2)cc1', '[Br-].[Br-].c1ccc2c(c1)ccc[n+]2CCCCCCCCCCCC[n+]1cccc2ccccc21', '[Br-].[Br-].c1ccc2c[n+](CCCCCCCCC[n+]3ccc4ccccc4c3)ccc2c1', '[Br-].[Br-].c1ccc2c(c1)ccc[n+]2CCCCCCCCCCC[n+]1cccc2ccccc21', 'C[N+](C)(C)CCCCCCCCCC[N+](C)(C)C.[Br-].[Br-]', '[Br-].[Br-].c1ccc2c(c1)ccc[n+]2CCCCCCCCC[n+]1cccc2ccccc21', '[Br-].[Br-].c1cc[n+](CCCCCCCCCCC[n+]2ccccc2)cc1', '[Br-].[Br-].c1cc[n+](CCCCCCCCCCC[n+]2ccccc2)cc1', '[Br-].[Br-].c1cc[n+](CCCCCCCCCCC[n+]2ccccc2)cc1', 'CN(C)Cc1ccc(CSCCNc2cc(Oc3ccc([N+](=O)[O-])cc3)c([N+](=O)[O-])cc2[N+](=O)[O-])o1', 'CN(C)Cc1ccc(CSCCNc2cc(Nc3ncccc3[N+](=O)[O-])c([N+](=O)[O-])cc2[N+](=O)[O-])o1', 'CN(C)Cc1ccc(CSCCNc2cc(N)c([N+](=O)[O-])cc2[N+](=O)[O-])o1', 'CN(C)Cc1ccc(CSCCNc2cc(Nc3ccccc3C#N)c([N+](=O)[O-])cc2[N+](=O)[O-])o1', 'CCCNc1cc(NCCSCc2ccc(CN(C)C)o2)c([N+](=O)[O-])cc1[N+](=O)[O-]', 'CNc1cc(NCCSCc2ccc(CN(C)C)o2)c([N+](=O)[O-])cc1[N+](=O)[O-]', 'CCCCCNc1cc(NCCSCc2ccc(CN(C)C)o2)c([N+](=O)[O-])cc1[N+](=O)[O-]', 'CN(C)Cc1ccc(CSCCNc2cc(Nc3ccc(F)cc3)c([N+](=O)[O-])cc2[N+](=O)[O-])o1', 'CN(C)Cc1ccc(CSCCNc2cc(NCc3ccccn3)c([N+](=O)[O-])cc2[N+](=O)[O-])o1', 'CN(C)Cc1ccc(CSCCNc2cc(NCCCN3CCOCC3)c([N+](=O)[O-])cc2[N+](=O)[O-])o1', 'CN(C)Cc1ccc(CSCCNc2cc(NCc3ccccc3)c([N+](=O)[O-])cc2[N+](=O)[O-])o1', 'CN(C)Cc1ccc(CSCCNc2cc(NCc3cccnc3)c([N+](=O)[O-])cc2[N+](=O)[O-])o1', 'CN(C)CCNc1cc(NCCSCc2ccc(CN(C)C)o2)c([N+](=O)[O-])cc1[N+](=O)[O-]', 'CN(C)Cc1ccc(CSCCNc2cc(Nc3ccccc3)c([N+](=O)[O-])cc2[N+](=O)[O-])o1', 'CC(C)=NNc1cc(NCCSCc2ccc(CN(C)C)o2)c([N+](=O)[O-])cc1[N+](=O)[O-]', 'CN(C)Cc1ccc(CSCCNc2cc(NN)c([N+](=O)[O-])cc2[N+](=O)[O-])o1', 'CN(C)Cc1ccc(CSCCNc2nnccc2[N+](=O)[O-])o1', 'CN(C)Cc1ccc(CSCCNc2ccc([N+](=O)[O-])cc2[N+](=O)[O-])o1', 'CN(C)Cc1ccc(CSCCNc2cc(F)c([N+](=O)[O-])cc2[N+](=O)[O-])o1', 'CN(C)Cc1ccc(CSCCNc2ccc(F)cc2[N+](=O)[O-])o1', 'CN(C)Cc1ccc(CSCCNc2cccc(F)c2C#N)o1', 'O=C1c2ccccc2C(=O)N1CCC1CCN(Cc2ccccc2)CC1', 'O=C1c2ccc([N+](=O)[O-])cc2C(=O)N1CC1CCN(Cc2ccccc2)CC1', 'O=C1c2ccc([N+](=O)[O-])cc2C(=O)N1CCC1CCN(Cc2ccccc2)CC1', 'O=C1c2ccc([N+](=O)[O-])cc2C(=O)N1C1CCN(Cc2ccccc2)CC1', 'Cc1c(C(C)C)c(=O)on1C(=O)N1CCC[C@H](C)C1', 'Cc1c(C)c2ccc(OCc3cccc(Cl)c3)cc2oc1=O', 'CC1=CC2Cc3nc4cccc(F)c4c(N)c3C(C1)C2', 'COc1cc(N)c(Cl)cc1C(=O)NC1CCN(Cc2ccccc2)CC1', 'COc1ccccc1CN1CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(Cc2ccccc2OC)Cc2ccc(cc2)-c2ccc(cc2)C1', 'CCCCCCCCCN(Cc1ccccc1OC)C(=O)CCCCCNCc1ccc(-c2ccc(-c3ccc(CNCCCCCC(=O)NCc4ccccc4OC)cc3)cc2)cc1', 'COc1ccccc1CN(CCCCCCCCN(Cc1ccccc1OC)C(=O)CCCCCN)C(=O)CCCCCN', 'COc1ccccc1CN1CCCCCCCCN(Cc2ccccc2OC)C(=O)CCCCCNCc2ccc(cc2)-c2ccc(cc2)CNCCCCCC1=O', 'COc1ccccc1CN1CCCCCCCCN(Cc2ccccc2OC)C(=O)CCCCCNCc2cccc(c2)Cc2cccc(c2)CNCCCCCC1=O', 'COc1ccccc1CN1CCCCCCCCN(Cc2ccccc2OC)C(=O)CCCCCNCc2ccc(cc2)CNCCCCCC1=O', 'CCCCC1=NC2(CCCC2)C(=O)N1Cc1ccc(-c2ccccc2C#N)cc1', 'CCCCC1=NC2(CCCC2)C(=O)N1Cc1ccc(-c2ccccc2C(=O)OC(C)(C)C)cc1', 'CCCCC1=NC2(CCCC2)C(=O)N1Cc1ccc(-c2ccccc2C(=O)OC)cc1', 'Cc1cccc([N+](=O)[O-])c1', 'Cc1cccc([N+](=O)[O-])c1', 'Nc1c2c(nc3ccccc13)CCCCCC2', 'Nc1c2c(nc3ccccc13)CCCCC2', 'COc1ccc2c(c1)CCCc1c-2nc2ccccc2c1N', 'CCCCC1(O)c2nc3ccccc3c(N)c2CC1C', 'CCCCC1=C(C)Cc2c1nc1ccccc1c2N', 'CC1=C(C)c2nc3ccccc3c(N)c2C1', 'Nc1c2c(nc3ccccc13)-c1ccccc1NCC2', 'Nc1c2c(nc3ccccc13)CCCc1ccccc1-2', 'COc1ccc2c(c1)CCc1c-2nc2ccccc2c1N', 'NC(=O)c1cc[n+](CCCC[n+]2ccc(/C=N/O)cc2)cc1.[Br-].[Br-]', 'Nc1c2c(nc3ccccc13)C1CCC2C1', 'Nc1c2c(nc3ccccc13)C1CCC2C1', 'Nc1c2c(nc3ccccc13)C1CCC2C1', 'Nc1c2c(nc3ccccc13)CCC2', 'CCCCCCCCCCCCCCCC[n+]1ccccc1.[Br-]', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)[C@@H]2N3C)cc1.O=C(O)C(O)C(O)C(=O)O', 'CCc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C.O=C(O)C(O)C(O)C(=O)O', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C.O=C(O)c1ccccc1O', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)O[C@@H]1N2C', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C)[C@@H]12.O=C(O)C(O)C(O)C(=O)O', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C)[C@H]2O1', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C.O=C(O)C(O)C(O)C(=O)O', 'C[N+](C)(C)CCCCCCCCCC[N+](C)(C)C', 'NC(=O)O', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CN(CCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1cccc(OC(=O)NCCCCCCCCN2CCOCC2)c1', 'CN(CCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1cccc(OC(=O)NCCCCCCCCN2CCOCC2)c1', 'CN(CCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1cccc(OC(=O)NCCCCCCCN2CCOCC2)c1', 'CN(CCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1cccc(OC(=O)NCCCCCCCN2CCOCC2)c1', 'CN(CCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1cccc(OC(=O)NCCCCCCN2CCOCC2)c1', 'CN(CCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1cccc(OC(=O)NCCCCCCN2CCOCC2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc(-c3ccccc3)cc2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc(-c3ccccc3)cc2)c1', 'CN(CCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1cccc(OC(=O)NCCCCCCCCCN2CCOCC2)c1', 'CN(CCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1cccc(OC(=O)NCCCCCCCCCN2CCOCC2)c1', 'CN(CCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1cccc(OC(=O)NCCCCCCCCCN2CCOCC2)c1', 'CN(CCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1cccc(OC(=O)NCCCCCCCCCN2CCOCC2)c1', 'CC(=O)N(CCC1CCN(Cc2ccccc2)CC1)c1ccccc1', 'O=C(c1ccncc1)N(CCC1CCN(Cc2ccccc2)CC1)c1ccccc1', 'O=C(c1ccc([N+](=O)[O-])cc1)N(CCC1CCN(Cc2ccccc2)CC1)c1ccccc1', 'O=C1c2ccc([N+](=O)[O-])cc2C(=O)N1CCC1CCN(Cc2ccc(Cl)cc2)CC1', 'CC(=O)N(C)CCC1CCN(Cc2ccccc2)CC1', 'CC(=O)N(CCC1CCN(Cc2ccccc2)CC1)c1ccc(F)cc1', 'O=C1c2ccccc2CN1CCC1CCN(Cc2ccccc2)CC1', 'COc1cccc(N(CCC2CCN(Cc3ccccc3)CC2)C(C)=O)c1', 'C(=C/C1CCN(Cc2ccccc2)CC1)\\c1noc2ccccc12', 'C(=C/C1CCN(Cc2ccccc2)CC1)\\c1noc2ccccc12', 'C(=C/C1CCN(Cc2ccccc2)CC1)\\c1noc2ccccc12', 'OC1CCN(Cc2ccc(OCCCN3CCCCC3)cc2)CC1', 'COc1ccc2c3c([nH]c2c1)C(C)=NCC3', 'c1cc(OCCCN2CCCCC2)ccc1CN1CCCCC1', 'CCCCCCCNC(=O)Oc1cccc(CN(C)CCCOc2ccc3ccc(=O)oc3c2)c1', 'CCCCCCCNC(=O)Oc1cccc(CN(CCCOc2ccc3c(=O)c4ccccc4oc3c2)C(C)C)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)cc(-c4ccccc4)oc3c2)c1', 'CCCCNC(=O)Oc1cccc(CN(C)CCCOc2ccc3ccc(=O)oc3c2)c1', 'CCCCNC(=O)Oc1cccc(CN(C)CCCOc2ccc3ccc(=O)oc3c2)c1', 'CCCCCCCNC(=O)Oc1cccc(CN(C)CCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3ccc(=O)oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3ccc(=O)oc3c2)c1', 'CCCCCCCNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)c4cccnc4oc3c2)c1', 'CCCCCCCNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)c4cccnc4oc3c2)c1', 'CCCNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)ccoc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)ccoc3c2)c1', 'c1ccc2c(NCCCCCCCCSc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCCCCCCCSc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCCCCCCCSc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'CCCCNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)cc(-c4ccccc4)oc3c2)c1', 'CN(CCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1cccc(OC(=O)Nc2ccccc2)c1', 'CN(CCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1cccc(OC(=O)Nc2ccccc2)c1', 'CN(CCCNc1c2c(nc3ccccc13)CCCC2)CCCNc1c2c(nc3ccccc13)CCCC2', 'CN(CCCNc1c2c(nc3ccccc13)CCCC2)CCCNc1c2c(nc3ccccc13)CCCC2', 'CCN(CCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1cccc(OC(=O)NC)c1', 'CCN(CCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1cccc(OC(=O)NC)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CCCCNC(=O)Oc1cccc(CN(CCCOc2ccc3c(=O)c4ccccc4oc3c2)C(C)C)c1', 'CCNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CCCN(CCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1cccc(OC(=O)NC)c1', 'CCCN(CCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1cccc(OC(=O)NC)c1', 'CCCCNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)c4cccnc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(CCCOc2ccc3c(=O)c4ccccc4oc3c2)C(C)C)c1', 'CCCNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CCCCNC(=O)Oc1cccc(CN(CCC)CCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CCCCNC(=O)Oc1cccc(CN(CCC)CCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CCN(CC)C(C)CN1c2ccccc2Sc2ccccc21', 'CCN(CC)C(C)CN1c2ccccc2Sc2ccccc21', 'CCN(CC)C(C)CN1c2ccccc2Sc2ccccc21', 'CCN(CC)C(C)CN1c2ccccc2Sc2ccccc21', 'CCCCNC(=O)Oc1cccc(CN(CC)CCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)c4cccnc4oc3c2)c1', 'O=C1/C(=C/c2ccc[nH]2)CN(Cc2ccccc2)C/C1=C\\c1ccc[nH]1', 'O=C1CCN(Cc2ccccc2)C/C1=C\\c1cccn1C(=O)c1ccccc1', 'O=C1CCN(Cc2ccccc2)C/C1=C\\c1cc[nH]c1', 'O=C1CCN(Cc2ccccc2)C/C1=C\\c1ccc[nH]1', 'CCCCNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'O=C1CCN(Cc2ccccc2)C/C1=C\\c1cn(C(=O)c2ccc(F)cc2)c2ccccc12', 'COc1ccc2c(c1)c(/C=C1\\CN(Cc3ccccc3)CCC1=O)cn2C(=O)c1ccccc1', 'O=C1CCN(Cc2ccccc2)C/C1=C\\c1c[nH]c2ccccc12', 'O=C1CCN(Cc2ccccc2)C/C1=C\\c1c[nH]c2ccc(OCc3ccccc3)cc12', 'COc1ccc2[nH]cc(/C=C3\\CN(Cc4ccccc4)CCC3=O)c2c1', 'N#Cc1ccc(C(=O)n2cc(/C=C3\\CN(Cc4ccccc4)CCC3=O)c3ccccc32)cc1', 'CC(=O)n1cc(/C=C2\\CN(Cc3ccccc3)CCC2=O)c2ccccc21', 'O=C1CCN(Cc2ccccc2)C/C1=C\\c1cn(C(=O)c2ccc([N+](=O)[O-])cc2)c2ccccc12', 'CC(C)(C)OC(=O)n1cc(/C=C2\\CN(Cc3ccccc3)CCC2=O)c2ccccc21', 'O=C1CCN(Cc2ccccc2)C/C1=C\\c1cn(C(=O)c2ccccc2)c2ccccc12', 'Cc1[nH]c(C)c(/C=C2\\CN(Cc3ccccc3)CCC2=O)c1C=O', 'O=C1/C(=C/c2cc[nH]c2)CN(Cc2ccccc2)C/C1=C\\c1cc[nH]c1', 'O=C1CCN(Cc2ccccc2)C/C1=C\\c1ccn(C(=O)c2ccccc2)c1', 'COc1ccc(C(=O)n2cc(/C=C3\\CN(Cc4ccccc4)CCC3=O)c3ccccc32)cc1', 'O=C1CCN(Cc2ccccc2)C/C1=C\\c1cn(C(=O)c2ccc(Cl)cc2)c2ccccc12', 'CCc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CCc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'Cc1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)[C@@H]2N3C)cc1', 'CCc1cccc(CC)c1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CCc1cccc(CC)c1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'Cc1cccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)[C@@H]2N3C)c1C', 'Cc1cccc(C)c1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'Cc1ccc(C)c(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)[C@@H]2N3C)c1', 'Cc1ccc(C)c(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)[C@@H]2N3C)c1', 'CN(C(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C)c1ccccc1', 'CN(C(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C)c1ccccc1', 'Cc1cccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)[C@@H]2N3C)c1', 'CC(C)c1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CC(C)c1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CN(C)Cc1cc2c(cc1OC(=O)Nc1ccccc1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CN(C)Cc1cc2c(cc1OC(=O)Nc1ccccc1)[C@]1(C)CCN(C)[C@@H]1N2C', 'Cc1cc(C)c(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)[C@@H]2N3C)c(C)c1', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'Cc1cc2c(cc1OC(=O)Nc1ccccc1)[C@]1(C)CCN(C)[C@@H]1N2C', 'Cc1cc(C)cc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)[C@@H]2N3C)c1', 'Cc1cc(C)cc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)[C@@H]2N3C)c1', 'Cc1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)[C@@H]2N3C)cc1C', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)[C@@H]2N3C)cc1', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)[C@@H]2N3C)cc1', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)[C@@H]2N3C)cc1', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)[C@@H]2N3C)cc1', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)[C@@H]2N3C)cc1', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)[C@@H]2N3C)cc1', 'Cc1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)[C@@H]2N3C)c(C)c1', 'CSc1nc(-c2ccc(Cl)cc2)nn1C(=O)N(C)C', 'COc1ccc(-c2nc(SC)n(C(=O)N(C)C)n2)cc1', 'CCCCCCSc1nc(-c2ccc(C)cc2)nn1C(=O)N(C)C', 'CSc1nc(-c2ccc(Cl)cc2)nn1C(=O)N(C)c1ccccc1', 'O=C(N1CCCCC1)n1nc(-c2ccc(Cl)cc2)nc1SCC1CC1', 'O=C(N1CCCCC1)n1nc(-c2ccc(Cl)cc2)nc1SCC(F)(F)F', 'CSc1nc(-c2ccc(Cl)cc2)nn1C(=O)N1CCCCC1', 'CSc1nc(-c2ccc(C)cc2)nn1C(=O)N(C)c1ccccc1', 'CSc1nc(-c2ccc(OC(F)(F)F)cc2)nn1C(=O)N(C)C', 'CSc1nc(-c2ccc(-c3ccccc3)cc2)nn1C(=O)N(C)C', 'Cc1ccc(-c2nc(SCC(F)(F)F)n(C(=O)N(C)C)n2)cc1', 'CCSc1nc(-c2ccc(Cl)cc2)nn1C(=O)N(C)C', 'CCCSc1nc(-c2ccc(Cl)cc2)nn1C(=O)N(C)C', 'CSc1nc(-c2ccc(C(F)(F)F)cc2)nn1C(=O)N(C)C', 'CCOc1nn(-c2cccc(OCc3ccccc3)c2)c(=O)o1', 'CSc1nc(-c2ccc3ccccc3c2)nn1C(=O)N(C)C', 'CCSc1nc(-c2ccc(OC)cc2)nn1C(=O)N(C)c1ccccc1', 'CSc1nc(-c2ccc(C)cc2)nn1C(=O)N(C)C', 'Nc1c2c(nc3cnccc13)OCCC2', 'Nc1c2c(nc3cnccc13)OCCC2', 'Nc1c2c(nc3cnccc13)CCCC2', 'Nc1c2c(nc3cccc(Cl)c13)OCCC2', 'Nc1c2c(nc3cccc(Cl)c13)CCCC2', 'Nc1c2c(nc3cccc(Cl)c13)CCCC2', 'Nc1c2c(nc3cccc(Cl)c13)CCCC2', 'Nc1c2c(nc3cccc(F)c13)CCCC2', 'Nc1c2c(nc3cccc(F)c13)CCCC2', 'Nc1c2c(nc3ccccc13)OCCC2', 'Nc1c2c(nc3ccccc13)OCCC2', 'COc1cccc2nc3c(c(N)c12)CCCO3', 'Nc1c2c(nc3ncccc13)OCCC2', 'Nc1c2c(nc3ncccc13)OCCC2', 'Nc1c2c(nc3ncccc13)CCCC2', 'COc1cccc2nc3c(c(N)c12)CCCC3', 'COc1cccc2nc3c(c(N)c12)CCCC3', 'Cc1cccc2nc3c(c(N)c12)CCCO3', 'Cc1cccc2nc3c(c(N)c12)CCCC3', 'Nc1c2c(nc3c1CCCO3)CCCC2', 'Nc1c2c(nc3cccc(F)c13)OCCC2', 'Nc1c2c(nc3ccc(Cl)cc13)OCCC2', 'Nc1c2c(nc3ccc(Cl)cc13)OCCC2', 'Nc1c2c(nc3c1CCCC3)CCCC2', 'Nc1c2c(nc3c1CCCO3)CCC2', 'COc1ccccc1CNCCCCCC(=O)NCCCCCCCCNC(=O)CCCCCNCc1ccccc1OC', 'COc1ccccc1CNCCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCNCc1ccccc1OC', 'COc1ccccc1CNCCCCCCNCCCCCNCCCCCCNCc1ccccc1OC', 'COc1ccccc1CNCCCCCCNCCCCCCCNCCCCCCNCc1ccccc1OC', 'O=C(O)CCCC[C@@H]1CCSS1', 'CCC(=O)Oc1cccc([N+](C)(C)C)c1', 'CC(=O)Oc1ccc(C)c([N+](C)(C)C)c1', 'C=CC[N+](C)(C)c1ccc(CCC(=O)CCc2ccc([N+](C)(C)CC=C)cc2)cc1', 'C=CC[N+](C)(C)c1ccc(CCC(=O)CCc2ccc([N+](C)(C)CC=C)cc2)cc1', 'CCC(=O)Oc1ccc(C)c([N+](C)(C)C)c1', 'CCC(=O)Oc1cc([N+](C)(C)C)ccc1C', 'CC(=O)Oc1cc([N+](C)(C)C)ccc1C', 'CC(=O)Oc1cccc([N+](C)(C)C)c1C', 'CC(=O)Oc1cccc([N+](C)(C)C)c1', 'Cc1ccc([N+](C)(C)C)cc1O', 'Cc1ccc(O)cc1[N+](C)(C)C', 'C[N+](C)(C)CCCBr', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CC[N+](C)(C)C1N2C', 'C[C@@H]([C@H]1CC[C@H]2[C@@H]3CC[C@H]4C(=O)C(NC(=O)c5ccccc5)=CC[C@]4(C)[C@H]3CC[C@]12C)N(C)C', 'C[C@@H]([C@H]1CC[C@H]2[C@@H]3CC[C@H]4C[C@@H](NC(=O)c5ccccc5)CC[C@]4(C)[C@H]3CC[C@]12C)N(C)C', 'CC(=O)N(C)[C@@H](C)[C@H]1CC[C@H]2[C@@H]3CC=C4C[C@@H](N(C)C)CC[C@]4(C)[C@H]3CC[C@]12C', 'CC(=O)CCC[N+](C)(C)C', 'C/C(=C\\C(=O)N(C)[C@H]1CC[C@@]2(C)[C@@H](CC[C@@H]3[C@@H]2CC[C@]2(C)C([C@H](C)N(C)C)=CC[C@@H]32)C1)C(C)C', 'C[N+](C)(C)c1cccc(O)c1', 'C/C(=C\\C(=O)N(C)[C@H]1CC[C@@]2(C)[C@@H](CC[C@@H]3[C@@H]2CC[C@]2(C)[C@@H]([C@H](C)N(C)C)CC[C@@H]32)C1)C(C)C', 'C[C@@H]([C@H]1CC[C@H]2[C@@H]3CC[C@H]4C[C@@H](N(C)C(=O)/C=C/c5ccccc5)CC[C@]4(C)[C@H]3CC[C@]12C)N(C)C', 'C/C=C(\\C)C(=O)N[C@@H]1[C@H](OC(C)=O)[C@@H]2CC[C@H]3C4=CC[C@H]([C@H](C)N(C)C)[C@@]4(C)CC[C@@H]3[C@@]2(C)C[C@@H]1O', 'C[C@@H]([C@H]1CC[C@H]2[C@@H]3CC[C@H]4C[C@@H](NC(=O)c5ccccc5)[C@@H](O)C[C@]4(C)[C@H]3CC[C@]12C)N(C)C', 'C/C=C(\\C)C(=O)N[C@H]1CC[C@@]2(C)C(=CC[C@H]3[C@@H]4CC[C@H]([C@H](C)N(C)C)[C@@]4(C)CC[C@@H]32)C1=O', 'CN[C@H]1CC[C@@]2(C)C(CC[C@H]3[C@@H]4CC[C@H]([C@H](C)N(C)C)[C@@]4(C)CC[C@@H]32)C1', 'C/C=C(\\C)C(=O)N[C@H]1CC[C@@]2(C)C(=CC[C@H]3C4=CC[C@H]([C@H](C)N(C)C)[C@@]4(C)CC[C@@H]32)C1=O', 'C/C=C(\\C)C(=O)N[C@H]1C(OC(C)=O)C[C@@]2(C)C(CC[C@H]3C4CC=C([C@H](C)N(C)C)[C@@]4(C)CC[C@@H]32)[C@H]1OC(C)=O', 'C/C=C(\\C)C(=O)N[C@@H]1C=C2CC[C@H]3[C@@H]4CC[C@H]([C@H](C)N(C)C)[C@@]4(C)CC[C@@H]3[C@@]2(C)C[C@@H]1O', 'CN[C@H]1CC[C@@]2(C)[C@@H](CC[C@@H]3[C@@H]2CC[C@]2(C)C([C@H](C)N(C)C)=CC[C@@H]32)C1', 'CN[C@@H](C)C1=CC[C@H]2[C@@H]3CC=C4C[C@@H](OC)CC[C@]4(C)[C@H]3CC[C@]12C', 'CC(=O)N[C@H]1CC[C@@]2(C)[C@@H](CC[C@@H]3[C@@H]2CC[C@]2(C)[C@@H]([C@H](C)N(C)C)CC[C@@H]32)C1', 'C/C=C(\\C)C(=O)N[C@@H]1C=C2CC[C@H]3C4=CC[C@H]([C@H](C)N(C)C)[C@@]4(C)CC[C@@H]3[C@@]2(C)CC1', 'C/C=C(\\C)C(=O)N[C@@H]1[C@H](O)C2CC[C@H]3C4=CC[C@H]([C@H](C)N(C)C)[C@@]4(C)CC[C@@H]3[C@@]2(C)C[C@@H]1O', 'CO[C@H]1CC[C@@]2(C)C(=CC[C@H]3[C@@H]4CC[C@H]([C@H](C)N(C)C)[C@@]4(C)CC[C@@H]32)C1', 'C[C@@H]([C@H]1CC[C@H]2[C@@H]3CC=C4C[C@@H](N(C)C)CC[C@]4(C)[C@H]3CC[C@]12C)N(C)C', 'CC(=O)N(C)[C@H]1CC[C@@]2(C)[C@@H](CC[C@@H]3[C@@H]2CC[C@]2(C)[C@@H]([C@H](C)N(C)C)CC[C@@H]32)C1', 'CO[C@H]1CC[C@@]2(C)C(=CC[C@@H]3[C@@H]2CC[C@]2(C)C([C@H](C)N(C)C=O)=CC[C@@H]32)C1', 'CC(=O)N(C)[C@@H](C)[C@H]1CC[C@H]2[C@@H]3CC[C@H]4C[C@@H](N(C)C)CC[C@]4(C)[C@H]3CC[C@]12C', 'C/C=C(\\C)C(=O)NC1=CC[C@@]2(C)[C@@H](CC[C@H]3C4CC[C@H]([C@H](C)N(C)C)[C@@]4(C)CC[C@@H]32)C1=O', 'CC(=O)O[C@@H]1C2CC[C@H]3C4CC=C([C@H](C)N(C)C)[C@@]4(C)CC[C@@H]3[C@@]2(C)CC[C@@H]1NC(=O)C=C(C)C', 'C[C@H](NC=O)[C@H]1CC[C@H]2[C@@H]3CC[C@H]4C[C@@H](N(C)C)CC[C@]4(C)[C@H]3CC[C@]12C', 'CN1CCC(CCN2C(=O)c3ccc([N+](=O)[O-])cc3C2=O)CC1', 'O=C1c2cccc([N+](=O)[O-])c2C(=O)N1CCC1CCN(Cc2ccccc2)CC1', 'O=C(c1ccccc1)c1ccc2c(c1)C(=O)N(CCC1CCN(Cc3ccccc3)CC1)C2=O', 'COc1ccc(CN2CCC(CCN3C(=O)c4ccc([N+](=O)[O-])cc4C3=O)CC2)cc1', 'O=C(C1CCCCC1)N(CCC1CCN(Cc2ccccc2)CC1)c1ccccc1', 'COc1ccc2c(c1)C(=O)N(CCC1CCN(Cc3ccccc3)CC1)C2=O', 'O=C(Nc1cccc2c1C(=O)N(CCC1CCN(Cc3ccccc3)CC1)C2=O)c1ccccc1', 'O=C(Nc1ccc2c(c1)C(=O)N(CCC1CCN(Cc3ccccc3)CC1)C2=O)c1ccccc1', 'CC(=O)Nc1ccc2c(c1)C(=O)N(CCC1CCN(Cc3ccccc3)CC1)C2=O', 'Nc1cccc2c1C(=O)N(CCC1CCN(Cc3ccccc3)CC1)C2=O', 'Nc1ccc2c(c1)C(=O)N(CCC1CCN(Cc3ccccc3)CC1)C2=O', 'O=c1[nH]c2cc(Cl)ccc2c(=O)n1CCC1CCN(Cc2ccccc2)CC1', 'O=C1CCCCN1CCC1CCN(Cc2ccccc2)CC1', 'O=c1c2ccccc2ccn1CCC1CCN(Cc2ccccc2)CC1', 'O=C1Nc2ccccc2CN1CCC1CCN(Cc2ccccc2)CC1', 'CC(=O)N(CCC1CCN(Cc2ccccc2)CC1)c1cccc(F)c1', 'COc1ccc(N(CCC2CCN(Cc3ccccc3)CC2)C(C)=O)cc1', 'CCC(=O)N(CCC1CCN(Cc2ccccc2)CC1)c1ccccc1', 'O=c1c2ccccc2ncn1CCC1CCN(Cc2ccccc2)CC1', 'O=C1Cc2ccccc2CN1CCC1CCN(Cc2ccccc2)CC1', 'O=c1[nH]c2ccccc2c(=O)n1CCC1CCN(Cc2ccccc2)CC1', 'O=C1c2ccccc2NCN1CCC1CCN(Cc2ccccc2)CC1', 'O=C1Cc2ccccc2C(=O)N1CCC1CCN(Cc2ccccc2)CC1', 'c1ccc(CN2CCC(CCN3CCc4ccccc4C3)CC2)cc1', 'O=C1c2ccccc2CCN1CCC1CCN(Cc2ccccc2)CC1', 'CC(=O)N(CCC1CCN(Cc2ccccc2)CC1)c1ccncc1', 'CCN(CCC1CCN(Cc2ccccc2)CC1)c1ccccc1', 'O=C1c2cccnc2C(=O)N1CCC1CCN(Cc2ccccc2)CC1', 'COc1ccc(C(=O)N(CCC2CCN(Cc3ccccc3)CC2)c2ccccc2)cc1', 'O=C(c1ccc(F)cc1)N(CCC1CCN(Cc2ccccc2)CC1)c1ccccc1', 'CCC1=CC2Cc3nc4cc(F)ccc4c(N)c3C(C1)C2', 'CCC1=CC2Cc3nc4cc(F)ccc4c(N)c3C(C1)C2', 'CCC1=CC2Cc3nc4cc(F)ccc4c(N)c3C(C1)C2', 'CCC1=CC2Cc3nc4cc(F)ccc4c(N)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(F)cc(F)c4c(N)c3C(C1)C2', 'CCC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2', 'CCC1=CC2Cc3nc4ccc(Cl)cc4c(N)c3C(C1)C2', 'CCC1=CC2Cc3nc4cccc(Cl)c4c(N)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2', 'CCC1=CC2Cc3nc4cccc(F)c4c(N)c3C(C1)C2', 'CCC1=CC2Cc3nc4cc(C)cc(C)c4c(N)c3C(C1)C2', 'CCC1=CC2Cc3nc4cc(F)cc(F)c4c(N)c3C(C1)C2', 'CCC1=CC2Cc3nc4cc(Cl)cc(Cl)c4c(N)c3C(C1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3[N+](=O)[O-])CC1)C2', 'Clc1ccc(CO/N=C(\\Cn2ccnc2)c2ccc(Cl)cc2Cl)c(Cl)c1', 'CC1CCN(CCc2ccccc2)CC1', 'CCCCCCNC(=O)Oc1ccc2c(c1)[C@@]1(C)CCN(CC)[C@H]1C2', 'Cn1cc[n+](COC[n+]2ccn(C)c2/C=N/O)c1/C=N/O', 'Cn1cc[n+](COCC2CCCc3ccccc32)c1/C=N/O.[Cl-]', 'CCC(OC[n+]1ccn(C)c1/C=N/O)C(C)(C)C.[Cl-]', 'CCCCCCOC[n+]1ccn(C)c1/C=N/O.[Cl-]', 'CC(CC(C)(C)C)OC[n+]1ccn(C)c1/C=N/O.[Cl-]', 'CC(OC[n+]1ccn(C)c1/C=N/O)C1CCCCC1.[Cl-]', 'Cn1cc[n+](COCC2CCCCC2)c1/C=N/O.[Cl-]', 'Cc1n(C)cc[n+]1COC(C)C(C)(C)C.[Cl-]', 'C#CC(C)C(C)OC[n+]1ccn(C)c1/C=N/O.[Cl-]', 'C#CCCCOC[n+]1ccn(C)c1/C=N/O.[Cl-]', 'C#CCC(CCC)OC[n+]1ccn(C)c1/C=N/O.[Cl-]', 'C#CC(CC)OC[n+]1ccn(C)c1/C=N/O.[Cl-]', 'C#CCCOC[n+]1ccn(C)c1/C=N/O.[Cl-]', 'C#CC(CCC)OC[n+]1ccn(C)c1/C=N/O.[Cl-]', 'CC#CCCOC[n+]1ccn(C)c1/C=N/O.[Cl-]', 'C#CC(C)(C)OC[n+]1ccn(C)c1/C=N/O.[Cl-]', 'C#CC(C)OC[n+]1ccn(C)c1/C=N/O.[Cl-]', 'C#CCCOC[n+]1ccc(/C=N/O)cc1.[Cl-]', 'C#CCCCC[n+]1cn(C)c(/C=N/O)n1.[Cl-]', 'C#CCCCC[n+]1ccn(C(C)(C)C)c1/C=N/O.[Cl-]', 'C#CC1(OC[n+]2ccn(C)c2/C=N/O)CCCC1.[Cl-]', 'C#CC1CCCCC1OC[n+]1ccn(C)c1/C=N/O.[Cl-]', 'C#CCC(C)OC[n+]1ccn(C)c1/C=N/O.[Cl-]', 'C#CCOC[n+]1ccn(C)c1/C=N/O.[Cl-]', 'C#CCCCC[n+]1ccccc1/C=N/O.[Cl-]', 'C#CCCCC[n+]1ccn(C)c1/C=N/O.[Cl-]', 'C#CCCOC[n+]1ccn(C)c1C.[Cl-]', 'C#CCN(C)C(C)Cc1cccc(OC(=O)N(C)C2CCCCC2)c1', 'C#CCN(C)C(C)Cc1cccc(OC(=O)N(C)CCCCCC)c1', 'C#CCN(C)C(C)Cc1cccc(OC(=O)N(C)CCCCCC)c1', 'C#CCN(C)C(C)Cc1cccc(OC(=O)N(C)CCCCCC)c1', 'C#CCN(C)C(C)Cc1cccc(OC(=O)N(C)CCCCCC)c1', 'C#CCN(C)C(C)Cc1cccc(OC(=O)N(C)CCCCCC)c1', 'C#CCN(C)C(C)Cc1cccc(OC(=O)N(C)CCC)c1', 'C#CC(C)NC1CCc2ccc(SC(=O)N(C)C)cc21', 'C#CCNC1CCc2ccc(OC(=O)N(C)C)cc2C1', 'C#CCC1CCCc2ccc(OC(=O)N(C)CC)cc21', 'C#CCC1CCCc2ccc(OC(=O)N(C)C)cc21', 'CCN(C)C(=O)Oc1ccc2c(c1)CC(N)CC2', 'CN(C)CCc1cccc(OC(=O)N(C)C)c1', 'CCN(C)C(=O)Oc1cccc(CCNC)c1', 'CC(c1cccc(OC(=O)N(C)C)c1)N(C)C', 'C#CCNC(C)Cc1cccc(OC(=O)N(C)CCC)c1', 'C#CCNC(C)Cc1cccc(OC(=O)N(C)CC)c1', 'C#CCNC(C)Cc1cccc(OC(=O)N(C)C)c1', 'C#CCN(C)CCc1cccc(OC(=O)N(C)CC)c1', 'CCN(C)C(=O)Oc1cccc(CCN)c1', 'CN(C)C(=O)Oc1ccc2c(c1)CC(N)CC2', 'CCN(C)C(=O)Oc1ccc2c(c1)C(N)CCC2', 'CN(C)C(=O)Oc1ccc2c(c1)C(N)CCC2', 'CN(C)C(=O)Oc1cccc(CCN)c1', 'CNCCc1cccc(OC(=O)N(C)C)c1', 'C#CCNC(C)Cc1cccc(OC(=O)N(C)C2CCCCC2)c1', 'C#CCN(C)C(C)Cc1cccc(OC(=O)N(C)C)c1', 'C#CCNC(C)Cc1cccc(OC(=O)N(C)CCCC)c1', 'CCCN(C)C(=O)Oc1cccc(CCN)c1', 'CCN(C)C(=O)Oc1cccc(C(C)N(C)C)c1', 'C#CCNCCc1cccc(OC(=O)N(C)C)c1', 'C#CC(C)NCCc1cccc(OC(=O)N(C)CC)c1', 'C#CCN(C)CCc1cccc(OC(=O)N(C)C)c1', 'C#CCNCCc1cccc(OC(=O)N(C)CCC)c1', 'C#CCNCCc1cccc(OC(=O)N(C)CC)c1', 'C#CC(C)NCCc1cccc(OC(=O)N(C)C)c1', 'Cc1cc(-c2ccc(Cl)cc2)nnc1NCCN1CCOCC1', 'CN(C)Cc1ccc(CNCCN2CCNC2=C(C#N)C#N)o1', 'N#CC(C#N)=C1NCCN1CCNCc1ccc(CN2CCCCC2)o1', 'CC1CCN(Cc2ccc(CNCCN3CCNC3=C(C#N)C#N)o2)CC1', 'N#CC(C#N)=C1NCCN1CCNCc1ccc(CN2CCC(O)CC2)o1', 'CC1CCCCN1Cc1ccc(CNCCN2CCNC2=C(C#N)C#N)o1', 'Oc1nc2ccccc2n1CCNCc1ccc(CN2CCCCC2)o1', 'CC1CCCC(C)N1Cc1ccc(CNCCN2CCNC2=C(C#N)C#N)o1', 'COc1cc(N)c(Cl)cc1C(=O)NCCNCc1ccc(CN(C)C)o1', 'N#CC(C#N)=C1NCCN1CCNCc1ccc(CN2CCOCC2)o1', 'N#CC(C#N)=C1NCCN1CCNCc1ccc(CN2CCCC(O)C2)o1', 'CNC(NCCNCc1ccc(CN(C(C)C)C(C)C)o1)=C(C#N)C#N', 'CNC(NCCNCc1ccc(CN(C)C)o1)=C(C#N)C#N', 'CC1CCCN(Cc2ccc(CNCCN3CCNC3=C(C#N)C#N)o2)C1', 'C/N=c1\\c(O)c(O)\\c1=N/CCNCc1ccc(CN2CCCCC2)o1', 'CNC(NCCNCc1ccc(CN2CCCC2)o1)=C(C#N)C#N', 'CN1CCN(Cc2ccc(CNCCN3CCNC3=C(C#N)C#N)o2)CC1', 'N#CC(C#N)=C1NCCN1CCNCc1ccc(CN2CCCC2)o1', 'N#CC(C#N)=C1NCCN1CCNCc1ccc(CN2CCSCC2)o1', 'CC1Cc2ccccc2C1=O', 'CC1Cc2ccccc2C1=O', 'Cc1cccc(F)c1', 'Cc1cccc(F)c1', 'CC1Cc2ccc(F)cc2C1=O', 'CC1Cc2ccc(F)cc2C1=O', 'COc1ccc(C)cc1', 'COc1ccc(C)cc1', 'COc1cccc(C)c1', 'COc1cccc(C)c1', 'Cc1ccc([N+](=O)[O-])cc1', 'Cc1ccc([N+](=O)[O-])cc1', 'COc1cc2c(cc1O)CC(C)C2=O', 'COc1cc2c(cc1O)CC(C)C2=O', 'Cc1ccc(F)cc1', 'Cc1ccc(F)cc1', 'CCOc1cc2c(cc1OCC)C(=O)C(C)C2', 'CCOc1cc2c(cc1OCC)C(=O)C(C)C2', 'Cc1cccc2c1CC(C)C2=O', 'Cc1cccc2c1CC(C)C2=O', 'COc1cc2c(cc1O)C(=O)C(C)C2', 'COc1cc2c(cc1O)C(=O)C(C)C2', 'COc1cc2c(cc1OC)C(=O)C(C)C2', 'COc1cc2c(cc1OC)C(=O)C(C)C2', 'N#CC(C#N)=C1Nc2ccccc2N1CCNCc1ccc(CN2CCCCC2)o1', 'CCN(C)Cc1ccc(CNCCN2CCNC2=C(C#N)C#N)o1', 'COc1cc(C)cc(OC)c1OC', 'COc1cc(C)cc(OC)c1OC', 'COc1ccc(OC)c2c1CC(C)C2=O', 'COc1ccc(OC)c2c1CC(C)C2=O', 'CC1Cc2cc3c(cc2C1=O)OCO3', 'CC1Cc2cc3c(cc2C1=O)OCO3', 'COc1cc2c(cc1OC(C)C)CC(C)C2=O', 'COc1cc2c(cc1OC(C)C)CC(C)C2=O', 'COc1ccc2c(c1OC)C(=O)C(C)C2', 'COc1ccc2c(c1OC)C(=O)C(C)C2', 'COc1ccc2c(c1)C(=O)C(C)C2', 'COc1ccc2c(c1)C(=O)C(C)C2', 'Cc1ccccc1F', 'Cc1ccccc1F', 'COc1ccc2c(c1OC)CC(C)C2=O', 'COc1ccc2c(c1OC)CC(C)C2=O', 'COc1cc2c(c(OC)c1OC)C(=O)C(C)C2', 'COc1cc2c(c(OC)c1OC)C(=O)C(C)C2', 'COc1cccc2c1CC(C)C2=O', 'COc1cccc2c1CC(C)C2=O', 'Cc1ccc2c(c1)C(=O)C(C)C2', 'Cc1ccc2c(c1)C(=O)C(C)C2', 'COc1cc2c(c(OC)c1)C(=O)C(C)C2', 'COc1cc2c(c(OC)c1)C(=O)C(C)C2', 'C=CC(=O)N1C/C(=C\\c2ccc(F)cc2)C(=O)/C(=C/c2ccc(F)cc2)C1', 'C=CC(=O)N1C/C(=C\\c2ccc(F)cc2)C(=O)/C(=C/c2ccc(F)cc2)C1', 'C=CC(=O)N1C/C(=C\\c2ccc(Cl)cc2)C(=O)/C(=C/c2ccc(Cl)cc2)C1', 'C=CC(=O)N1C/C(=C\\c2ccc(Cl)cc2)C(=O)/C(=C/c2ccc(Cl)cc2)C1', 'CC[N+](C)(CC)CCC[n+]1c(-c2ccccc2)c2cc(N)ccc2c2ccc(N)cc21.[I-].[I-]', 'CC[N+](C)(CC)CCC[n+]1c(-c2ccccc2)c2cc(N)ccc2c2ccc(N)cc21.[I-].[I-]', 'CC[N+](C)(CC)CCC[n+]1c(-c2ccccc2)c2cc(N)ccc2c2ccc(N)cc21.[I-].[I-]', 'CC[N+](C)(CC)CCC[n+]1c(-c2ccccc2)c2cc(N)ccc2c2ccc(N)cc21.[I-].[I-]', 'CC[N+](C)(CC)CCC[n+]1c(-c2ccccc2)c2cc(N)ccc2c2ccc(N)cc21.[I-].[I-]', 'CC[N+](C)(CC)CCC[n+]1c(-c2ccccc2)c2cc(N)ccc2c2ccc(N)cc21.[I-].[I-]', 'CC[N+](C)(CC)CCC[n+]1c(-c2ccccc2)c2cc(N)ccc2c2ccc(N)cc21.[I-].[I-]', 'CC[N+](C)(CC)CCC[n+]1c(-c2ccccc2)c2cc(N)ccc2c2ccc(N)cc21.[I-].[I-]', 'CC[N+](C)(CC)CCC[n+]1c(-c2ccccc2)c2cc(N)ccc2c2ccc(N)cc21.[I-].[I-]', 'CC[N+](C)(CC)CCC[n+]1c(-c2ccccc2)c2cc(N)ccc2c2ccc(N)cc21.[I-].[I-]', 'CC[N+](C)(CC)CCC[n+]1c(-c2ccccc2)c2cc(N)ccc2c2ccc(N)cc21.[I-].[I-]', 'CC[N+](C)(CC)CCC[n+]1c(-c2ccccc2)c2cc(N)ccc2c2ccc(N)cc21.[I-].[I-]', 'CC[N+](C)(CC)CCC[n+]1c(-c2ccccc2)c2cc(N)ccc2c2ccc(N)cc21.[I-].[I-]', 'CC[N+](C)(CC)CCC[n+]1c(-c2ccccc2)c2cc(N)ccc2c2ccc(N)cc21.[I-].[I-]', 'C=CC(=O)N1C/C(=C\\c2ccccc2)C(=O)/C(=C/c2ccccc2)C1', 'C=CC(=O)N1C/C(=C\\c2ccccc2)C(=O)/C(=C/c2ccccc2)C1', 'CC1CCN(CC2Cc3ccccc3C2=O)CC1', 'CN1CCC(CCCC2Cc3ccccc3C2=O)CC1', 'CN1CCC(CC2=Cc3ccccc3C2)CC1', 'CC1CCN(C(=O)c2ccccc2)CC1', 'Cc1ccccc1CN1CCC(C)CC1', 'CC1CCN(Cc2ccccc2)CC1', 'CN1CCC(CC/C=C2\\Cc3ccccc3C2=O)CC1', 'CN1CCN(CC2Cc3ccccc3C2=O)CC1', 'CN1CCC(C2Cc3ccccc3C2=O)CC1', 'CN1CCC(=C2Cc3ccccc3C2=O)CC1', 'CN1CCC(/C=C2\\Cc3ccccc3C2=O)CC1', 'CN1CCC(/C=C/C=C2\\Cc3ccccc3C2=O)CC1', 'CN1CCC(C[C@H]2Cc3ccccc3C2=O)CC1', 'CC(C)N1CCN(c2ccc(OC[C@H]3CO[C@](Cn4cncn4)(c4ccc(Cl)cc4Cl)O3)cc2)CC1', 'CC1CCN(Cc2ccccc2[N+](=O)[O-])CC1', 'CC1CCN(CC2CCCCC2)CC1', 'CN1CCC(CCC2C(=O)Nc3ccccc3C2=O)CC1', 'CN1CCC(CCN2C(=O)c3ccccc3C2=O)CC1', 'CC1CCN(Cc2cccc3ccccc23)CC1', 'CC1CCN(Cc2ccc3ccccc3c2)CC1', 'CN1CCC(C[C@@H]2Cc3ccccc3C2=O)CC1', 'C/N=C1/CCc2c1n(C)c1ccc(OC(=O)NC)c(Cl)c21', 'C/N=C1/CCc2c1n(C)c1ccc(OC(=O)NC)cc21', 'C/N=C1/CCc2c1n(C)c1ccc(OC(=O)NC)c(Br)c21', 'CNC(=O)Oc1ccc2c(c1)c1c(n2C)/C(=N\\CCc2ccccc2)CC1', 'CCC/N=C1/CCc2c1n(C)c1ccc(OC(=O)NC)cc21', 'CC/N=C1/CCc2c1n(C)c1ccc(OC(=O)NC)c(Cl)c21', 'CC/N=C1/CCc2c1n(C)c1ccc(OC(=O)NC)c(Br)c21', 'C/N=C1/CCc2c1n(C)c1ccc(OC(=O)N(C)C)c(Br)c21', 'C/N=C1/CCc2c1n(C)c1ccc(OC(=O)N(C)C)c(Br)c21', 'CCCCNC(=O)Oc1ccc2c(c1Br)c1c(n2C)/C(=N\\C)CC1', 'C/N=C1/CCc2c1n(C)c1ccc(OC(=O)NCc3ccccc3)c(Br)c21', 'C/N=C1/CCc2c1n(C)c1cc(Cl)c(OC(=O)NC)cc21', 'CC/N=C1/CCc2c1n(C)c1ccc(OC(=O)NC)cc21', 'CCCn1c2c(c3c(Br)c(OC(=O)NC)ccc31)CC/C2=N/CC', 'Clc1ccc2c(c1)N=C1CCCCCN1C2', 'N=c1c2ccc(Cl)cc2nc2n1CCCCC2', 'Clc1cccc2c1N=C1CCCCCN1C2', 'Clc1ccc2c(c1)CN1CCCCCC1=N2', 'c1ccc2c(c1)CN1CCCC1=N2', 'c1ccc2c(c1)CN1CCCC1=N2', 'c1ccc2c(c1)CN1CCCCCC1=N2', 'c1ccc2c(c1)CN1CCCCCCCC1=N2', 'Clc1cccc2c1CN1CCCCCC1=N2', 'c1ccc2c(c1)CN1CCCCCCC1=N2', 'c1ccc2c(c1)CN1CCCCC1=N2', 'O=c1ccc2cc3ccoc3cc2o1', 'Nc1c2c(nc3c(Cl)cccc13)CCCC2', 'Nc1c2c(nc3c(Cl)cccc13)CCCC2', 'Nc1c2c(nc3cccc(Cl)c13)C1CCC2C1', 'Nc1c2c(nc3cccc(Cl)c13)C1CCC2C1', 'Nc1c2c(nc3ccc(Cl)cc13)C1CCC2C1', 'Nc1c2c(nc3ccc(Cl)cc13)C1CCC2C1', 'Nc1c2c(nc3c(Cl)cccc13)C1CCC2C1', 'Nc1c2c(nc3c(Cl)cccc13)C1CCC2C1', 'Nc1c2c(nc3cc(Cl)ccc13)C1CCC2C1', 'Nc1c2c(nc3cc(Cl)ccc13)C1CCC2C1', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccc(C)cc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(CC3CCCCC3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccc([N+](=O)[O-])cc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3cccc([N+](=O)[O-])c3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)/C(=C/C1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)/C(=C/C1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3cccc(C)c3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3C)CC1)C2', 'CC[N+](C)(CCCCCCOCCCCCCCCOCCCCCC[N+](C)(CC)Cc1ccccc1OC)Cc1ccccc1OC', 'CCN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)Cc1ccccc1OC)Cc1ccccc1OC', 'CCN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)Cc1ccccc1OC)Cc1ccccc1OC', 'CCN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)Cc1ccccc1OC)Cc1ccccc1OC', 'CCN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)Cc1ccccc1OC)Cc1ccccc1OC', 'CCN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)Cc1ccccc1OC)Cc1ccccc1OC', 'CCN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)Cc1ccccc1OC)Cc1ccccc1OC', 'COc1cccc(CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(C)Cc2cccc(OC)c2)c1', 'CCN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)Cc1ccccc1Cl)Cc1ccccc1Cl', 'CCN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)Cc1ccccc1C(F)(F)F)Cc1ccccc1C(F)(F)F', 'COc1ccc(CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(C)Cc2ccc(OC)cc2)cc1', 'CCN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)Cc1ccccc1)Cc1ccccc1', 'CC[N+](C)(CCCCCC(=O)N(CCCCCCCCN(C(=O)CCCCC[N+](C)(CC)Cc1ccccc1OC)C(C)C)C(C)C)Cc1ccccc1OC', 'COc1ccccc1CN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(Cc1ccccc1OC)C(C)C)C(C)C', 'CCN(CCCCCCOCCCCCCCCOCCCCCCN(CC)Cc1ccccc1OC)Cc1ccccc1OC', 'CC[N+](C)(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCC[N+](C)(CC)Cc1ccccc1OC)Cc1ccccc1OC', 'CC[N+](C)(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCC[N+](C)(CC)Cc1ccccc1OC)Cc1ccccc1OC', 'CC[N+](C)(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCC[N+](C)(CC)Cc1ccccc1OC)Cc1ccccc1OC', 'CC[N+](C)(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCC[N+](C)(CC)Cc1ccccc1OC)Cc1ccccc1OC', 'CCCOc1ccccc1C[N+](C)(CC)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCC[N+](C)(CC)Cc1ccccc1OCCC', 'CC[N+](C)(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCC[N+](C)(CC)Cc1ccccc1C)Cc1ccccc1C', 'CCN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)Cc1ccccc1[N+](=O)[O-])Cc1ccccc1[N+](=O)[O-]', 'CCN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)Cc1ccccc1[N+](=O)[O-])Cc1ccccc1[N+](=O)[O-]', 'CCN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)Cc1ccccc1[N+](=O)[O-])Cc1ccccc1[N+](=O)[O-]', 'CCN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)Cc1ccccc1[N+](=O)[O-])Cc1ccccc1[N+](=O)[O-]', 'CCN(CCCCCC(=O)N(CCCCCCCCN(C(=O)CCCCCN(CC)Cc1ccccc1OC)C(C)C)C(C)C)Cc1ccccc1OC', 'CC[N+](C)(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCC[N+](C)(CC)Cc1ccccc1)Cc1ccccc1', 'CC[N+](C)(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCC[N+](C)(CC)Cc1ccccc1Cl)Cc1ccccc1Cl', 'CC[N+](C)(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCC[N+](C)(CC)Cc1ccccc1C(F)(F)F)Cc1ccccc1C(F)(F)F', 'CCCOc1ccccc1CN(CC)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)Cc1ccccc1OCCC', 'CCN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)Cc1ccccc1C)Cc1ccccc1C', 'CC(=O)N(CCCNc1c2c(nc3ccccc13)CCCC2)CCCSc1c2c(nc3ccccc13)CCCC2', 'CN(CCCCSc1c2c(nc3ccccc13)CCCC2)CCCNc1c2c(nc3ccccc13)CCCC2', 'CC(=O)N(CCCCSc1c2c(nc3ccccc13)CCCC2)CCCNc1c2c(nc3ccccc13)CCCC2', 'CCNC(=O)Oc1ccc2c(c1)C1(C)CCOC1O2', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCOC1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCOC1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCOC1N2C', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)C1(C)CCOC1O2', 'CCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCOC1N2C', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCOC1N2C', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)C2(C)CCOC2O3)cc1', 'CCNC(=O)Oc1ccc2c(c1)C1(C)COC(C1)O2', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCOC2N3C)cc1', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)C1(C)COC(C1)O2', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)C2(C)COC(C2)O3)cc1', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)C2(C)COC(C2)O3)cc1', 'CCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2C', 'Nc1ccc2c(c1)c(-c1ccccc1)[n+](CCCNCCCNCCCNc1c3c(nc4ccccc14)CCCC3)c1cc(N)ccc21', 'Nc1ccc2c(c1)c(-c1ccccc1)[n+](CCCNCCCNCCCNc1c3c(nc4ccccc14)CCCC3)c1cc(N)ccc21', 'Nc1ccc2c(c1)c(-c1ccccc1)[n+](CCCNCCCNCCCNc1c3c(nc4ccccc14)CCCC3)c1cc(N)ccc21', 'COc1ccccc1CN(C)CCCN(C)CCCNCCC[n+]1c(-c2ccccc2)c2cc(N)ccc2c2ccc(N)cc21', 'COc1ccccc1CN(C)CCCN(C)CCCNCCC[n+]1c(-c2ccccc2)c2cc(N)ccc2c2ccc(N)cc21', 'COc1cc2c(cc1OC)C(=O)C(Oc1ccc(CN(C)C)cc1)C2', 'COc1cc2c(cc1OC)C(=O)C(Oc1cccc(C(C)N(C)C)c1)C2', 'COc1cc2c(cc1OC)C(=O)C(Oc1cccc(CN(C)C)c1)C2', 'CCN(CC)Cc1ccc(OC2Cc3cc(OC)c(OC)cc3C2=O)cc1', 'CCN(CC)Cc1cccc(OC2Cc3cc(OC)c(OC)cc3C2=O)c1', 'COc1cc2c(cc1OC)C(=O)C(Oc1ccc(C(C)N(C)C)cc1)C2', 'CCN(CC)C(C)c1ccc(OC2Cc3cc(OC)c(OC)cc3C2=O)cc1', 'COc1cc2c(cc1OC)C(=O)C(Oc1cccc(CN3CCN(C)CC3)c1)C2', 'CCN(CC)C(C)c1cccc(OC2Cc3cc(OC)c(OC)cc3C2=O)c1', 'COc1cc2c(cc1OC)C(=O)C(Oc1cccc(CN3CCCC3)c1)C2', 'COc1cc2c(cc1OC)C(=O)C(Oc1ccc(CN3CCOCC3)cc1)C2', 'COc1cc2c(cc1OC)C(=O)C(Oc1cccc(CN3CCCCC3)c1)C2', 'COc1cc2c(cc1OC)C(=O)C(Oc1ccc(CN3CCN(C)CC3)cc1)C2', 'COc1cc2c(cc1OC)C(=O)C(Oc1cccc(CN3CCOCC3)c1)C2', 'COc1cc2c(cc1OC)C(=O)C(Oc1ccc(CN3CCCC3)cc1)C2', 'COc1cc2c(cc1OC)C(=O)C(Oc1ccc(CN3CCCC3)cc1)C2', 'COc1cc2c(cc1OC)C(=O)C(Oc1cccc(C(C)N3CCN(C)CC3)c1)C2', 'COc1cc2c(cc1OC)C(=O)C(Oc1ccc(CN3CCCCC3)cc1)C2', 'COc1cc2c(cc1OC)C(=O)C(Oc1cccc(C(C)N3CCCCC3)c1)C2', 'COc1cc2c(cc1OC)C(=O)C(Oc1ccc(C(C)N3CCCC3)cc1)C2', 'COc1cc2c(cc1OC)C(=O)C(Oc1ccc(C(C)N3CCCCC3)cc1)C2', 'COc1cc2c(cc1OC)C(=O)C(Oc1cccc(C(C)N3CCOCC3)c1)C2', 'COc1cc2c(cc1OC)C(=O)C(Oc1ccc(C(C)N3CCOCC3)cc1)C2', 'COc1cc2c(cc1OC)C(=O)C(Oc1cccc(C(C)N3CCCC3)c1)C2', 'COc1cc2c(cc1OC)C(=O)C(Oc1ccc(C(C)N3CCN(C)CC3)cc1)C2', 'COc1ccc(-c2cc3c(c(=O)o2)C[C@@]2(O)[C@@]4(C)C(=O)[C@@]5(OC)C(=O)O[C@]4(CC[C@@]2(C)O3)C5(C)C)cc1', 'COc1ccc(-c2cc3c(c(=O)o2)C[C@@]2(O)[C@@]4(C)C=CC(=O)C(C)(C)[C@@]4(O)CC[C@@]2(C)O3)cc1', 'COc1ccc(-c2cc3c(c(=O)o2)C[C@@]2(O)[C@@]4(C)CCC(=O)C(C)(C)[C@@]4(O)CC[C@@]2(C)O3)cc1', 'COC1=CC(=O)[C@]2(C)[C@]3(O)Cc4c(cc(-c5cc(OC)c(OC)c(OC)c5)oc4=O)O[C@]3(C)CC[C@]2(O)C1(C)C', 'COc1ccc(-c2cc3c(c(=O)o2)C[C@@]2(O)[C@@]4(C)CCC(=O)OC(C)(C)[C@@]4(O)CC[C@@]2(C)O3)cc1', 'COc1ccc(-c2cc3c(c(=O)o2)C[C@@]2(O)[C@@]4(C)CCC(=O)OC(C)(C)[C@@]4(O)CC[C@@]2(C)O3)cc1', 'CCCCNC(=O)OCCO', 'CCCCNC(=O)OCCCCCO', 'CCCCNC(=O)OCCCCCCCCO', 'CCCCNC(=O)OCCOC(=O)NCCCC', 'CCCCNC(=O)OCCCCCOC(=O)NCCCC', 'CCCCNC(=O)OCCCCCCCCOC(=O)NCCCC', 'CCCCNC(=O)OCCCCCCCCCCOC(=O)NCCCC', 'CCCCNC(=O)OCCCCCCCCCCCCOC(=O)NCCCC', 'CCCCNC(=O)OCCCCCCCCCCCCCCOC(=O)NCCCC', 'CCCCNC(=O)OCCCCCCCCCCCCCCCCOC(=O)NCCCC', 'CCCCNC(=O)OCCCCCCCCCCO', 'CCCCNC(=O)OCCCCCCCCCCCCO', 'CCCCNC(=O)OCCCCCCCCCCCCCCO', 'CCCCNC(=O)OCCCCCCCCCCCCCCCCO', 'CC[N+](CC)(CC)CCOc1cccc(OCC[N+](CC)(CC)CC)c1OCC[N+](CC)(CC)CC', 'CC1(C)C=Cc2c(ccc(C(=O)/C=C/c3ccc(O)cc3)c2O)O1', 'CC1(C)C=Cc2c(ccc(C(=O)/C=C/c3ccc(O)cc3)c2O)O1', 'CC1(C)C=Cc2c(ccc(C(=O)/C=C/c3ccc(O)cc3)c2O)O1', 'COc1ccc(/C=C/C(=O)c2ccc3c(c2O)C=CC(C)(C)O3)cc1', 'Cc1ccc(S(=O)(=O)N(Cc2ccccc2)c2ccc(N)cc2O)cc1', 'Cc1nn(S(=O)(=O)c2cc(C(F)(F)F)cc(C(F)(F)F)c2)cc1-c1nsc(Nc2ccccc2)n1', 'Cc1cc(C2(c3cc(C)c(O)c(C)c3)CCCCC2)cc(C)c1O', 'Cc1c(/N=C/c2ccccc2O)cc2c(c1N)C(C)(C)CC2(C)C', 'CC(Cc1ccc(O)cc1)C1CCC(C)C(c2ccc(O)cc2)C1', 'CN(C)c1ccc(-c2c3cc(O)c(=O)cc-3oc3cc(O)c(O)cc23)cc1', 'Fc1ccc(N(Cc2ccc3c(c2)OCO3)C(=S)Nc2cccc(C(F)(F)F)c2)cc1', 'O=C(CC1S/C(=N/c2ccc(C3CCCCC3)cc2)NC1=O)Nc1ccc(OC(F)(F)F)cc1', 'O=C(CC1S/C(=N/c2ccc(C3CCCCC3)cc2)NC1=O)Nc1ccc(OC(F)(F)F)cc1', 'Oc1ccc(C2CC(c3ccccc3O)=Nc3ccccc3S2)cc1', 'CC(C)Cc1ccc(C(=O)CC(Sc2ccccc2N)c2cccc([N+](=O)[O-])c2)cc1', 'Cc1nc(N)ncc1-c1nnnn1-c1ccc(-c2ccccc2)cc1', 'COc1ccc(-c2cc(-c3ccc(N(C)C)cc3)nc(NS(=O)(=O)c3ccc(NC(C)=O)cc3)n2)cc1', 'COc1ccc(/N=C/c2cn(-c3ccc(OC)cc3)cc2/C=N/c2ccc(OC)cc2)cc1', 'COc1cc2onc(CCC3CCN(Cc4ccccc4)CC3)c2cc1C', 'COc1cc2onc(CCC3CCN(Cc4ccccc4)CC3)c2cc1C', 'COc1cc2cc(C(=O)C3CC[N+](C)(Cc4ccccc4)CCC3O)sc2cc1OC', 'COc1cc2cc(C(=O)C3CC[N+](C)(Cc4ccccc4)CCC3O)sc2cc1OC', 'COc1cc2cc(C(=O)C3CC[N+](C)(Cc4ccccc4)CCC3OC)sc2cc1OC', 'COc1cc2cc(C(=O)C3CC[N+](C)(Cc4ccccc4)CCC3OC)sc2cc1OC', 'COc1cc2c(cc1OC)SC(C(=O)CCc1cc[n+](CC3CCC3)cc1)C2', 'COc1cc2c(cc1OC)SC(C(=O)CCc1cc[n+](CC3CCC3)cc1)C2', 'CCOCC[n+]1ccc(CCC(=O)C2Cc3cc(OC)c(OC)cc3S2)cc1', 'CCOCC[n+]1ccc(CCC(=O)C2Cc3cc(OC)c(OC)cc3S2)cc1', 'COc1cc2c(cc1OC)SC(C(=O)CCc1cc[n+](Cc3ccsc3)cc1)C2', 'COc1cc2c(cc1OC)SC(C(=O)CCc1cc[n+](Cc3ccsc3)cc1)C2', 'COc1cc2c(cc1OC)SC(C(=O)CCc1cc[n+](Cc3ccccc3)cc1)C2', 'COc1cc2c(cc1OC)SC(C(=O)CCc1cc[n+](Cc3ccccc3)cc1)C2', 'COc1cc2c(cc1OC)SC(C(=O)CC1C=CC[N+](C)(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)SC(C(=O)CC1C=CC[N+](C)(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)SC(CC1CC[N+](C)(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)SC(CC1CC[N+](C)(Cc3ccccc3)CC1)C2', 'O=C(CCc1cc[n+](Cc2ccccc2)cc1)c1ccc2c(c1)OCO2', 'O=C(CCc1cc[n+](Cc2ccccc2)cc1)c1ccc2c(c1)OCO2', 'COc1cc2c(cc1OC)C(=O)C(C1CC[N+](C)(Cc3ccccc3)CC1)=C2', 'COc1cc2c(cc1OC)C(=O)C(C1CC[N+](C)(Cc3ccccc3)CC1)=C2', 'COc1cc2cc(C(=O)C3CCC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC', 'COc1cc2cc(C(=O)C3CCC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC', 'COc1cc2cc(C(=O)CC3CCCC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC', 'COc1cc2cc(C(=O)CC3CCCC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC', 'COC(=O)c1ccc(C[n+]2ccc(CCC(=O)C3Cc4cc(OC)c(OC)cc4S3)cc2)cc1', 'COC(=O)c1ccc(C[n+]2ccc(CCC(=O)C3Cc4cc(OC)c(OC)cc4S3)cc2)cc1', 'CCC[n+]1ccc(CCC(=O)C2Cc3cc(OC)c(OC)cc3S2)cc1', 'CCC[n+]1ccc(CCC(=O)C2Cc3cc(OC)c(OC)cc3S2)cc1', 'C=CC[n+]1ccc(CCC(=O)C2Cc3cc(OC)c(OC)cc3S2)cc1', 'C=CC[n+]1ccc(CCC(=O)C2Cc3cc(OC)c(OC)cc3S2)cc1', 'COc1cc2c(cc1OC)SC(C(=O)CCc1cc[n+](CC3CC3)cc1)C2', 'COc1cc2c(cc1OC)SC(C(=O)CCc1cc[n+](CC3CC3)cc1)C2', 'COCC[n+]1ccc(CCC(=O)C2Cc3cc(OC)c(OC)cc3S2)cc1', 'COCC[n+]1ccc(CCC(=O)C2Cc3cc(OC)c(OC)cc3S2)cc1', 'COc1cc2c(cc1OC)SC(C(=O)CC1(O)CC[N+](C)(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)SC(C(=O)CC1(O)CC[N+](C)(Cc3ccccc3)CC1)C2', 'COc1cc2cc(C(=O)C3=CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC', 'COc1cc2cc(C(=O)C3=CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC', 'COc1cc2cc(C3CC[N+](C)(Cc4ccccc4)CC3O)sc2cc1OC', 'COc1cc2cc(C3CC[N+](C)(Cc4ccccc4)CC3O)sc2cc1OC', 'CN1C(=O)Cc2cc3onc(CCC4CCN(Cc5ccccc5)CC4)c3cc21', 'CN1C(=O)Cc2cc3onc(CCC4CCN(Cc5ccccc5)CC4)c3cc21', 'COc1cc2cc(C(=O)CC3CCC(O)C[N+](C)(Cc4ccccc4)CC3)sc2cc1OC', 'COc1cc2cc(C(=O)CC3CCC(O)C[N+](C)(Cc4ccccc4)CC3)sc2cc1OC', '[Br-].[Br-].c1cc(C[n+]2ccc(N3CCCCC3)cc2)cc(-c2cccc(C[n+]3ccc(N4CCCCC4)cc3)c2)c1', 'COc1cc2c(cc1OC)SC(C(=O)CCc1cc[n+](Cc3ccc(F)cc3)cc1)C2', 'COc1cc2c(cc1OC)SC(C(=O)CCc1cc[n+](Cc3ccc(F)cc3)cc1)C2', 'COC(=O)C[N+]1(C)CCC(CCC(=O)C2Cc3cc(OC)c(OC)cc3S2)CC1', 'COC(=O)C[N+]1(C)CCC(CCC(=O)C2Cc3cc(OC)c(OC)cc3S2)CC1', 'COc1cc2cc(C(=O)C3CCC[N+](C)(CCOc4ccccc4)CC3)sc2cc1OC', 'COc1cc2cc(C(=O)C3CCC[N+](C)(CCOc4ccccc4)CC3)sc2cc1OC', 'COc1cc2cc(C(=O)C3CCC[N+](C)(CC#N)CC3)sc2cc1OC', 'COc1cc2cc(C(=O)C3CCC[N+](C)(CC#N)CC3)sc2cc1OC', 'COc1cc2cc(C3=CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC', 'COc1cc2cc(C3=CC[N+](C)(Cc4ccccc4)CC3)sc2cc1OC', 'COCC[N+]1(C)CCCC(C(=O)c2cc3cc(OC)c(OC)cc3s2)CC1', 'COCC[N+]1(C)CCCC(C(=O)c2cc3cc(OC)c(OC)cc3s2)CC1', 'COc1cc2cc(C(=O)C3CCC[N+](C)(CC(=O)OC(C)(C)C)CC3)sc2cc1OC', 'COc1cc2cc(C(=O)C3CCC[N+](C)(CC(=O)OC(C)(C)C)CC3)sc2cc1OC', 'C=[N+]1N=C(CCC2CCN(Cc3ccccc3)CC2)c2ccccc21', 'C=[N+]1N=C(CCC2CCN(Cc3ccccc3)CC2)c2ccccc21', 'COc1cc2cc(C(=O)C3CCC[N+](C)(CC4CCCCO4)CC3)sc2cc1OC', 'COc1cc2cc(C(=O)C3CCC[N+](C)(CC4CCCCO4)CC3)sc2cc1OC', '[Br-].[Br-].c1cc(C[n+]2ccc(N3CCCCC3)cc2)ccc1C[n+]1ccc(N2CCCCC2)cc1', 'COc1cc2c(cc1OC)SC(C(=O)CCc1cc[n+](C)cc1)C2', 'COc1cc2c(cc1OC)SC(C(=O)CCc1cc[n+](C)cc1)C2', 'CC[n+]1ccc(CCC(=O)C2Cc3cc(OC)c(OC)cc3S2)cc1', 'CC[n+]1ccc(CCC(=O)C2Cc3cc(OC)c(OC)cc3S2)cc1', 'COc1cc2c(cc1OC)SC(C(=O)CC(O)C1CC[N+](C)(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)SC(C(=O)CC(O)C1CC[N+](C)(Cc3ccccc3)CC1)C2', 'C[C@H]1COC2=C1C(=O)C(=O)c1c2ccc2c1CCCC2(C)C', 'CNC(=O)Oc1cccc(C(C)N(C)C)c1.Cl', 'CNC(=O)Oc1cccc2c1OCC1CCN(C)C21', 'CNC(=O)Oc1cccc2c1CCC1CCN(C)C21', 'CCCCCCCNC(=O)Oc1cccc2c1OCC1CCN(C)C21', 'CCNC(=O)Oc1cccc2c1OCC1CCN(C)C21', 'CNC(=O)Oc1cccc2c1SCC1CCN(C)C21', 'CC(NC(=O)Oc1cccc2c1OCC1CCN(C)C21)c1ccccc1', 'CCNC(=O)Oc1cccc2c1CCC1CCN(C)C21', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)O/C(=C\\c2ccccc2)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)O/C(=C\\c2ccccc2)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)O/C(=C\\c2cc(Cl)cc(Cl)c2)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)O/C(=C\\c2cc(Cl)cc(Cl)c2)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)C/C(=C\\c2cc(OC)c(OC)c(OC)c2)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)C/C(=C\\c2cc(OC)c(OC)c(OC)c2)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(c2)O/C(=C\\c2cccc4ccccc24)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(c2)O/C(=C\\c2cccc4ccccc24)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(c2)O/C(=C\\c2ccc4ccccc4c2)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(c2)O/C(=C\\c2ccc4ccccc4c2)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)O/C(=C\\c2c4ccccc4cc4ccccc24)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)O/C(=C\\c2c4ccccc4cc4ccccc24)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)O/C(=C\\c2ccc(Cl)c(Cl)c2)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)O/C(=C\\c2ccc(Cl)c(Cl)c2)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)O/C(=C\\c2ccc(Cl)c(Cl)c2)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)O/C(=C\\c2cc(OC)c(OC)c(OC)c2)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)O/C(=C\\c2cccc4ccccc24)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)O/C(=C\\c2cccc4ccccc24)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)O/C(=C\\c2cccc4ccccc24)C3=O)c1', 'CCC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)O/C(=C\\c2ccc4ccccc4c2)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)OC/C(=C\\c2cc(OC)c(OC)c(OC)c2)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)CC/C(=C\\c2cc(OC)c(OC)c(OC)c2)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(c2)O/C(=C\\c2cc(OC)c(OC)c(OC)c2)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)OCC3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)OCC3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CC(C)N(CCN1CC(=O)N(CCc2ccc(Cl)cc2Cl)CC1=O)C(C)C', 'COc1ccc(CCN2CC(=O)N(CCc3ccccn3)CC2=O)cc1OC', 'CC1CCCCN1CCCN1CC(=O)N(CCc2ccc(F)cc2)CC1=O', 'O=C1CN(CCN2CCOCC2)C(=O)CN1CCC(c1ccccc1)c1ccccc1', 'CCN(CC)CCCN1CC(=O)N(C2CC2)CC1=O', 'CCN(CC)CCN1CC(=O)N(CCc2ccccc2)CC1=O', 'O=C1CN(CCN2CCCC2)C(=O)CN1CCCn1ccnc1', 'O=C1C(=O)C(Cl)=C(Cl)C(Cl)=C1Cl', 'O=C1C(Cl)=C(Cl)C(=O)C(Cl)=C1Cl', 'Cc1c(Cl)nc2ccccc2c1C(=O)NCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'Cc1c(N2CCN(C)CC2)nc2ccccc2c1C(=O)NCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'Cc1c(N2CCN(C)CC2)nc2ccccc2c1C(=O)NCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'Cc1c(N2CCN(C)CC2)nc2ccccc2c1C(=O)NCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(CCCCC1CCSS1)NCCCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(CCCCC1CCSS1)NCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCCCC1CCSS1)NCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCCCC1CCSS1)NCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCCCC1CCSS1)NCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCCCC1CCSS1)NCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCCCC1CCSS1)NCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCCCC1CCSS1)NCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCCCC1CCSS1)NCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Nc1c2c(nc3cc(Cl)ccc13)CC(CNC(=O)CCCCC1CCSS1)CC2', 'O=C(CCCCC1CCSS1)NCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(CCCCC1CCSS1)NCCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(CCCCC1CCSS1)NCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(CCCCC1CCSS1)NCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(CCCCC1CCSS1)NCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'NCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'NCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(Nc1ccccc1)OC1CN2CCC1CC2', 'COc1cc2ccnc3c2c(c1OC)COC3(CC(O)CO)OC', 'C=CCC1(OC)OCc2c(OC)c(OC)cc3ccnc1c23', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCNC2', 'COc1cc2ccnc3c2c(c1OC)C(Br)OC3=O', 'COc1cc2ccnc3c2c(c1OC)COC3=O', 'CCCCCCCCc1ccc(O)cc1', 'CN[C@@H](CCOc1ccc([N+](=O)[O-])cc1)c1ccc(OC(=O)N(C)C)cc1', 'CN[C@@H](CCOc1ccc([N+](=O)[O-])cc1)c1ccc(OC(=O)N(C)C)cc1', 'CN(C)C(=O)Oc1ccc2c(c1)C=CCN(C)[C@@H]2CCOc1ccc([N+](=O)[O-])cc1', 'O=C(CCCCCNc1c2c(nc3ccccc13)CCCC2)NCCc1c[nH]c2ccc(O)cc12', 'O=C(CCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)NCCc1c[nH]c2ccccc12', 'O=C(CCCCCNc1c2c(nc3cccc(Cl)c13)CCCC2)NCCc1c[nH]c2ccccc12', 'O=C(CCCCCCNc1c2c(nc3cc(Cl)cc(Cl)c13)CCCC2)NCCc1c[nH]c2ccccc12', 'COc1ccc2[nH]cc(CCNC(=O)CCCCCCNc3c4c(nc5ccccc35)CCCC4)c2c1', 'COc1ccc2[nH]cc(CCNC(=O)CCCCCCNc3c4c(nc5cc(Cl)cc(Cl)c35)CCCC4)c2c1', 'O=C(CCCCCCNc1c2c(nc3ccccc13)CCCC2)NCCc1c[nH]c2ccccc12', 'COc1cc2c(cc1OC)-c1cc3ccc(OC)c(OC)c3c[n+]1CC2', 'CNC(=O)Oc1ccc2c(c1)[C@@]1(C)CCO[C@H]1N2C', 'CCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCO[C@@H]1N2C', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCO[C@@H]1N2C', 'CCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCO[C@@H]1O2', 'CCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCO[C@@H]1O2', 'CCNC(=O)Oc1ccc2c(c1)[C@@]1(C)CCO[C@H]1O2', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCO[C@@H]1O2', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCO[C@@H]1O2', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@@]1(C)CCO[C@H]1O2', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCO[C@@H]2O3)cc1', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@@]2(C)CCO[C@H]2O3)cc1', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@@]1(C)CCO[C@H]1N2C', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCO[C@@H]2N3C)cc1', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@@]1(C)CO[C@H](C1)O2', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CO[C@@H](C1)O2', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CO[C@@H](C2)O3)cc1', 'CCNC(=O)Oc1ccc2c(c1)[C@@]1(C)CO[C@H](C1)O2', 'CCNC(=O)Oc1ccc2c(c1)[C@]1(C)CO[C@@H](C1)O2', 'CNC(=O)Oc1ccc2c(c1)[C@@]1(C)CCN(C)[C@H]1N2C', 'CCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCO[C@@H]1N2C', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCO[C@@H]1N2C', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@@]1(C)CCN(C)[C@H]1N2C', 'COc1cc2ccnc(C(O)c3ccccc3OC)c2cc1OC', 'COc1ccc(C(O)c2nccc3cc(OC)c(OC)cc23)cc1OC', 'COc1ccc(CC2=[N+](C)CCc3cc(OC)c(OC)cc32)cc1OC', 'C/C=C1\\C2C=C(C)C[C@]1(N)CC(NCCCCCCCNc1c3c(nc4ccccc14)CCCC3)C2(C)C', 'C/C=C1\\C2C=C(C)C[C@]1(N)CC(NCCCCCCCNc1c3c(nc4ccccc14)CCCC3)C2', 'C/C=C1\\C2C=C(C)C[C@]1(C(=O)OC)CC(NCCCCCCCNc1c3c(nc4ccccc14)CCCC3)C2', 'CCC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2', 'CCC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2', 'CCC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2', 'CCC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2', 'CCC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2', 'CCC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2', 'CCC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2', 'Fc1ccc(C(c2ccc(F)cc2)N2CCN(CCN3CCOCC3)CC2)cc1', 'Fc1ccc(C(c2ccc(F)cc2)N2CCN(CCN3CCCCC3)CC2)cc1', 'Fc1ccc(C(c2ccc(F)cc2)N2CCN(CCN3CCCC3)CC2)cc1', 'Cc1nc2ccccn2c(=O)c1CCN1CCN(C(c2ccc(F)cc2)c2ccc(F)cc2)CC1', 'CC1=C(CCN2CCN(C(c3ccc(F)cc3)c3ccc(F)cc3)CC2)C(=O)C2CCCCC2=N1', 'Cc1cc2c(cc1CN1CCN(C(c3ccc(F)cc3)c3ccc(F)cc3)CC1)OCO2', 'COc1cc(Br)c(CN2CCN(C(c3ccc(F)cc3)c3ccc(F)cc3)CC2)cc1OC', 'N#Cc1ccccc1-c1ccc(CN2CCN(C(c3ccc(F)cc3)c3ccc(F)cc3)CC2)cc1', 'CN(C)C(=O)Oc1cccc([N+](C)(C)C)c1.COS(=O)(=O)[O-]', 'CN(C)C(=O)Oc1cccc([N+](C)(C)C)c1.COS(=O)(=O)[O-]', 'CN(C)C(=O)Oc1cccc([N+](C)(C)C)c1.COS(=O)(=O)[O-]', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1.O=C(O)C(O)C(O)C(=O)O', 'NC(=O)c1cc[n+](COC[n+]2ccccc2/C=N/O)cc1', 'NC(=O)c1cc[n+](COC[n+]2ccccc2/C=N/O)cc1', 'NC(=O)c1cc[n+](COC[n+]2ccccc2/C=N/O)cc1', 'C[n+]1ccccc1C=NO', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2NC=O', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2NS(C)(=O)=O', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2N', 'COc1cc(C(=O)NCCCCCCC(=O)NNc2c3c(nc4ccccc24)CCCC3)cc(OC)c1OC', 'COc1cc(C(=O)NCCCCCCCC(=O)NNc2c3c(nc4ccccc24)CCCC3)cc(OC)c1OC', 'COc1cc(C(=O)NCCCCCCCCC(=O)NNc2c3c(cc4ccccc24)CCCC3)cc(OC)c1OC', 'COc1cc(CCC(=O)NNc2c3c(nc4ccccc24)CCCC3)cc(OC)c1OC', 'COc1cc(CCC(=O)NCC(=O)NNc2c3c(nc4ccccc24)CCCC3)cc(OC)c1OC', 'COc1cc(CCC(=O)NCCC(=O)NNc2c3c(nc4ccccc24)CCCC3)cc(OC)c1OC', 'O=c1[nH]c(=O)c2nc[nH]c2[nH]1', 'COc1cc(C(=O)NNc2c3c(nc4ccccc24)CCCC3)cc(OC)c1OC', 'COc1cc(CCC(=O)NCCCC(=O)NNc2c3c(nc4ccccc24)CCCC3)cc(OC)c1OC', 'COc1cc(C(=O)NCC(=O)NNc2c3c(nc4ccccc24)CCCC3)cc(OC)c1OC', 'COc1cc(CCC(=O)NCCCCC(=O)NNc2c3c(nc4ccccc24)CCCC3)cc(OC)c1OC', 'COc1cc(C(=O)NCCC(=O)NNc2c3c(nc4ccccc24)CCCC3)cc(OC)c1OC', 'COc1cc(CCC(=O)NCCCCCC(=O)NNc2c3c(nc4ccccc24)CCCC3)cc(OC)c1OC', 'COc1cc(C(=O)NCCCC(=O)NNc2c3c(nc4ccccc24)CCCC3)cc(OC)c1OC', 'COc1cc(C(=O)NCCCCC(=O)NNc2c3c(nc4ccccc24)CCCC3)cc(OC)c1OC', 'COc1cc(C(=O)NCCCCCC(=O)NNc2c3c(nc4ccccc24)CCCC3)cc(OC)c1OC', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccccc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccccc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccccc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccc(F)cc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccc(F)cc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccc(F)cc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccccc1[N+](=O)[O-])CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccccc1[N+](=O)[O-])CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1cccc([N+](=O)[O-])c1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1cccc([N+](=O)[O-])c1)CCCC3', 'CC[C@H]1C[C@@H]2C[C@H]3c4[nH]c5cc([C@@H]6C[C@H]7C(C(=O)OC)[C@@H](Cc8c6[nH]c6ccccc86)N(C)C[C@H]7CC)ccc5c4CCN(C2)[C@@H]13', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccc([N+](=O)[O-])cc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccc([N+](=O)[O-])cc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccc(C)cc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccc(C)cc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccc(C)cc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccccc1OC)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccccc1OC)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1cccc(OC)c1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1cccc(OC)c1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccc(OC)cc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccc(OC)cc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccc(OC)cc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccc(OC)cc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccc(OC)cc1)CCCC3', 'COc1ccc2c(c1)c1ccnc(C)c1n2C', 'COc1ccc2c(c1)c1ccnc(C)c1n2C', 'COc1ccc2[nH]c3c[n+](C)ccc3c2c1', 'COc1ccc2[nH]c3c[n+](C)ccc3c2c1', 'COc1ccc2c(c1)c1cc[n+](C)cc1n2C', 'COc1ccc2c(c1)c1cc[n+](C)cc1n2C', 'COc1ccc2[nH]c3c(C)[n+](C)ccc3c2c1', 'COc1ccc2[nH]c3c(C)[n+](C)ccc3c2c1', 'C[n+]1ccc2c(c1)[nH]c1ccc(O)cc12', 'C[n+]1ccc2c(c1)[nH]c1ccc(O)cc12', 'Cn1c2ccc(O)cc2c2cc[n+](C)cc21', 'Cn1c2ccc(O)cc2c2cc[n+](C)cc21', 'Cc1c2[nH]c3ccc(O)cc3c2cc[n+]1C', 'Cc1c2[nH]c3ccc(O)cc3c2cc[n+]1C', 'Cc1c2c(cc[n+]1C)c1cc(O)ccc1n2C', 'Cc1c2c(cc[n+]1C)c1cc(O)ccc1n2C', 'CC(C)=CCC/C(C)=C/Cc1c(O)cc2c(c1O)C(=O)C[C@@H](c1ccc(O)cc1)O2', 'CC(C)=CCC/C(C)=C/Cc1c(O)cc2c(c1O)C(=O)C[C@@H](c1ccc(O)c(O)c1)O2', 'CC(C)=CCC/C(C)=C/Cc1c(O)cc2c(c1O)C(=O)C[C@@H](c1ccc(O)c(O)c1)O2', 'O=C1C(=O)N(Cc2ccc(Cc3ccc(CN4C(=O)C(=O)c5ccccc54)cc3)cc2)c2ccccc21', 'O=C1C(=O)N(Cc2ccc(Cl)c(Cl)c2)c2ccccc21', 'COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(C)Cc1ccccc1OC.O=C(O)C(=O)O', 'COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(C)Cc1ccccc1OC.O=C(O)C(=O)O', 'CCN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)Cc1ccccc1OC)Cc1ccccc1OC.O=C(O)C(=O)O', 'CCN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)Cc1ccccc1OC)Cc1ccccc1OC.O=C(O)C(=O)O', 'CCN(CCCCCC(=O)N1CCC(C2CCN(C(=O)CCCCCN(CC)Cc3ccccc3OC)CC2)CC1)Cc1ccccc1OC.O=C(O)C(=O)O', 'CCN(CCCCCC(=O)N1CCC(C2CCN(C(=O)CCCCCN(CC)Cc3ccccc3OC)CC2)CC1)Cc1ccccc1OC.O=C(O)C(=O)O', 'CCN(CCCCCC(=O)N1CCC(CC2CCN(C(=O)CCCCCN(CC)Cc3ccccc3OC)CC2)CC1)Cc1ccccc1OC.O=C(O)C(=O)O', 'CCN(CCCCCC(=O)N1CCC(CCC2CCN(C(=O)CCCCCN(CC)Cc3ccccc3OC)CC2)CC1)Cc1ccccc1OC.O=C(O)C(=O)O', 'CCN(CCCCCC(=O)N1CCC(CCCC2CCN(C(=O)CCCCCN(CC)Cc3ccccc3OC)CC2)CC1)Cc1ccccc1OC.O=C(O)C(=O)O', 'CCN(CCCCCC(=O)Nc1ccc(-c2ccc(NC(=O)CCCCCN(CC)Cc3ccccc3OC)cc2)cc1)Cc1ccccc1OC.O=C(O)C(=O)O', 'CCN(CCCCCC(=O)Nc1ccc(Cc2ccc(NC(=O)CCCCCN(CC)Cc3ccccc3OC)cc2)cc1)Cc1ccccc1OC.O=C(O)C(=O)O', 'CCN(CCCCCC(=O)Nc1ccc(CCc2ccc(NC(=O)CCCCCN(CC)Cc3ccccc3OC)cc2)cc1)Cc1ccccc1OC.O=C(O)C(=O)O', 'CC[N+](C)(CCCCCC(=O)N1CCC(CCC2CCN(C(=O)CCCCC[N+](C)(CC)Cc3ccccc3OC)CC2)CC1)Cc1ccccc1OC.[I-].[I-]', 'CC[N+](C)(CCCCCC(=O)N1CCC(CCCC2CCN(C(=O)CCCCC[N+](C)(CC)Cc3ccccc3OC)CC2)CC1)Cc1ccccc1OC.[I-].[I-]', 'CC[N+](C)(CCCCCC(=O)Nc1ccc(CCc2ccc(NC(=O)CCCCC[N+](C)(CC)Cc3ccccc3OC)cc2)cc1)Cc1ccccc1OC.[I-].[I-]', 'CCN(CCCCCCNc1ccc(-c2ccc(NCCCCCCN(CC)Cc3ccccc3OC)cc2)cc1)Cc1ccccc1OC', 'CCN(CCCCCCNc1ccc(Cc2ccc(NCCCCCCN(CC)Cc3ccccc3OC)cc2)cc1)Cc1ccccc1OC', 'N#Cc1ccc2[nH]cc(CCC(=O)NCCCCCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)c2c1', 'O=C(/C=C/c1c[nH]c2ccccc12)NCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(NCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)c1c[nH]c2ccccc12', 'O=C(NCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)c1c[nH]c2ccccc12', 'O=C(NCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)c1c[nH]c2ccccc12', 'O=C(NCCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)c1c[nH]c2ccccc12', 'O=C(Cc1c[nH]c2ccccc12)NCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(Cc1c[nH]c2ccccc12)NCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(Cc1c[nH]c2ccccc12)NCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(Cc1c[nH]c2ccccc12)NCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(Cc1c[nH]c2ccc(Br)cc12)NCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(Cc1c[nH]c2ccc(Br)cc12)NCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(Cc1c[nH]c2ccc(Br)cc12)NCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(Cc1c[nH]c2ccc(Br)cc12)NCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCCc1c[nH]c2ccccc12)NCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCCc1c[nH]c2ccccc12)NCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCCc1c[nH]c2ccccc12)NCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCCc1c[nH]c2ccccc12)NCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Cn1cc(C(=O)NCCCCCCCNc2c3c(nc4cc(Cl)ccc24)CCCC3)c2ccccc21', 'O=C(NCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)c1n[nH]c2ccccc12', 'O=C(CCc1c[nH]c2ccccc12)NCCCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(CCc1c[nH]c2ccccc12)NCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCc1c[nH]c2ccccc12)NCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCc1c[nH]c2ccccc12)NCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCc1c[nH]c2ccccc12)NCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCc1c[nH]c2ccccc12)NCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCc1c[nH]c2ccccc12)NCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCc1c[nH]c2ccccc12)NCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCc1c[nH]c2ccccc12)NCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCc1c[nH]c2ccccc12)NCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCc1c[nH]c2ccccc12)NCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCc1c[nH]c2ccccc12)NCCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'CN(CCCNC(=O)CCc1c[nH]c2ccccc12)CCCNc1c2c(nc3ccccc13)CCCC2', 'CN(CCCNC(=O)CCc1c[nH]c2ccccc12)CCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'CN(CCCCCCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1cccc(Cl)c1', 'CN(CCCCCCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1ccc(Cl)cc1', 'CN(CCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1ccccc1', 'CN(CCCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1ccccc1', 'CN(CCCCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1ccccc1', 'CN(CCCCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1ccccc1', 'COc1cccc(CN(C)CCCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1OC', 'CN(CCCCCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1ccccc1', 'CCN(CCCCCCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1ccccc1OC', 'CN(CCCCCCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1ccccc1', 'CN(CCCCCCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1ccccc1', 'CN(CCCCCCCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1ccccc1', 'COc1ccc2c(=O)c3ccc(OCCCCCCCN(C)Cc4ccccc4)cc3oc2c1', 'COc1cc2oc3cc(OCCCCCCCN(C)Cc4ccccc4)ccc3c(=O)c2cc1OC', 'COc1ccccc1CN(C)CCCOc1ccc2c(=O)c3ccccc3oc2c1', 'COc1ccccc1CN(C)CCCCCCCOc1ccc2c(=O)c3ccccc3oc2c1', 'COc1cccc(CN(C)CCCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'COc1cccc(CN(C)CCCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'COc1ccc(CN(C)CCCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)cc1', 'C[C@@H]1O[C@@H](OC[C@H]2O[C@@H](Oc3c(-c4ccc(O)c(O)c4)oc4cc(O)cc(O)c4c3=O)[C@H](O)[C@@H](O)[C@@H]2O)[C@H](O)[C@H](O)[C@H]1O', '[Br-].[Br-].c1cc[n+](CCCCCC[n+]2ccccc2)cc1', '[Br-].[Br-].c1cc[n+](CCCCCCCC[n+]2ccccc2)cc1', '[Br-].[Br-].c1cc[n+](CCCCCCCCCC[n+]2ccccc2)cc1', '[Br-].[Br-].c1ccc2c(c1)ccc[n+]2CCCCCCCCCC[n+]1cccc2ccccc21', '[Br-].[Br-].c1ccc2c(c1)ccc[n+]2CCCCCCCCCC[n+]1cccc2ccccc21', '[Br-].[Br-].c1ccc2c(c1)ccc[n+]2CCCCCCCCCC[n+]1cccc2ccccc21', 'C[C@H]1C=C2N3CC[C@@H](O)C24[C@H](CC(=O)[C@H]4C[C@H](O)C3)C1', 'CC1=C[C@H]2[C@H](O)C(=O)[C@H]3CCCN4CCC[C@H]2[C@]34C1', 'Cc1cccc2c3c(ccc12)C1=C(C(=O)C3=O)[C@@H](C)CO1', 'Cc1cccc2c3c(ccc12)C1=C(C(=O)C3=O)[C@@H](C)CO1', 'Cc1cccc2c3c(ccc12)C1=C(C(=O)C3=O)[C@@H](C)CO1', 'COc1cc2cc(-c3cccc(CN(C)Cc4ccccc4)c3)c(=O)oc2cc1OC', 'COc1ccc2cc(-c3cccc(CN(C)Cc4ccccc4)c3)c(=O)oc2c1', 'COc1ccc2cc(-c3ccc(CN(C)Cc4ccccc4)cc3)c(=O)oc2c1', 'COc1ccc2oc(=O)c(-c3cccc(CN(C)Cc4ccccc4)c3)cc2c1', 'COc1ccc2oc(=O)c(-c3ccc(CN(C)Cc4ccccc4)cc3)cc2c1', 'COc1cc2cc(-c3ccc(CN(C)Cc4ccccc4[N+](=O)[O-])cc3)c(=O)oc2cc1OC', 'COc1cc2cc(-c3ccc(CN(C)Cc4cccc([N+](=O)[O-])c4)cc3)c(=O)oc2cc1OC', 'COc1cc2cc(-c3ccc(CN(C)Cc4ccc([N+](=O)[O-])cc4)cc3)c(=O)oc2cc1OC', 'COc1cc2cc(-c3ccc(CN(C)Cc4ccccc4C)cc3)c(=O)oc2cc1OC', 'COc1cc2cc(-c3ccc(CN(C)Cc4cccc(C)c4)cc3)c(=O)oc2cc1OC', 'COc1cc2cc(-c3ccc(CN(C)Cc4ccc(C)cc4)cc3)c(=O)oc2cc1OC', 'COc1ccccc1CN(C)Cc1ccc(-c2cc3cc(OC)c(OC)cc3oc2=O)cc1', 'COc1cccc(CN(C)Cc2ccc(-c3cc4cc(OC)c(OC)cc4oc3=O)cc2)c1', 'COc1ccc(CN(C)Cc2ccc(-c3cc4cc(OC)c(OC)cc4oc3=O)cc2)cc1', 'CCN(Cc1ccccc1)Cc1ccc(-c2cc3cc(OC)c(OC)cc3oc2=O)cc1', 'CCN(Cc1ccccc1)Cc1ccc(-c2cc3cc(OC)c(OC)cc3oc2=O)cc1', 'CCN(Cc1ccccc1)Cc1ccc(-c2cc3cc(OC)c(OC)cc3oc2=O)cc1', 'COc1cc2cc(-c3ccc(CN(CCO)Cc4ccccc4)cc3)c(=O)oc2cc1OC', 'CN(Cc1ccccc1)Cc1ccc(-c2cc3cc([N+](=O)[O-])ccc3oc2=O)cc1', 'CN(Cc1ccccc1)Cc1ccc(-c2cc3cc(N)ccc3oc2=O)cc1', 'COc1cc2oc3cc(CN(C)Cc4ccccc4)ccc3c(=O)c2cc1OC', 'COc1cc2occ(-c3ccc(CN(C)Cc4ccccc4)cc3)c(=O)c2cc1OC', 'COc1cc2occ(-c3cccc(CN(C)Cc4ccccc4)c3)c(=O)c2cc1OC', 'CCN(Cc1ccccc1)Cc1cccc(-c2coc3cc(OC)c(OC)cc3c2=O)c1', 'COc1cc2oc(-c3cccc(CN(C)Cc4ccccc4)c3)cc(=O)c2cc1OC', 'COc1cc2oc(-c3ccc(CN(C)Cc4ccccc4)cc3)cc(=O)c2cc1OC', 'COc1ccc2oc(-c3ccc(CN(C)Cc4ccccc4)cc3)cc(=O)c2c1', 'COc1ccc2c(=O)cc(-c3ccc(CN(C)Cc4ccccc4)cc3)oc2c1', 'COc1cc2cc(-c3ccc(C[N+](C)(C)Cc4ccccc4)cc3)c(=O)oc2cc1OC.[I-]', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccccc2OC)=CC1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccccc2OC)=CC1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccccc2OC)=CC1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccccc2OC)=CC1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccccc2OC)=CC1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccccc2OC)=CC1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccccc2OC)=CC1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccccc2OC)=CC1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccccc2OC)=CC1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccccc2OC)=CC1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccccc2OC)=CC1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccccc2OC)=CC1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccccc2OC)=CC1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccccc2OC)=CC1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=C(F)C(=O)C(NCCCCCCN(CC)Cc2ccccc2OC)=C(F)C1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=C(Cl)C(=O)C(NCCCCCCN(CC)Cc2ccccc2OC)=C(Cl)C1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=C(Br)C(=O)C(NCCCCCCN(CC)Cc2ccccc2OC)=C(Br)C1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=C(C)C(=O)C(NCCCCCCN(CC)Cc2ccccc2OC)=C(C)C1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=C(c2ccccc2)C(=O)C(NCCCCCCN(CC)Cc2ccccc2OC)=C(c2ccccc2)C1=O)Cc1ccccc1OC', 'CCN(CCCCCCNc1ccc(NCCCCCCN(CC)Cc2ccccc2OC)cc1)Cc1ccccc1OC', 'CCN(CCNC1=CC(=O)C(NCCN(CC)Cc2ccccc2OC)=CC1=O)Cc1ccccc1OC', 'CCN(CCCNC1=CC(=O)C(NCCCN(CC)Cc2ccccc2OC)=CC1=O)Cc1ccccc1OC', 'CCN(CCCCNC1=CC(=O)C(NCCCCN(CC)Cc2ccccc2OC)=CC1=O)Cc1ccccc1OC', 'CCN(CCCCCNC1=CC(=O)C(NCCCCCN(CC)Cc2ccccc2OC)=CC1=O)Cc1ccccc1OC', 'CCN(CCCCCCCNC1=CC(=O)C(NCCCCCCCN(CC)Cc2ccccc2OC)=CC1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)CCc2ccccc2OC)=CC1=O)CCc1ccccc1OC', 'CCN(Cc1ccccc1OC)C(=O)CCCCCNC1=CC(=O)C(NCCCCCC(=O)N(CC)Cc2ccccc2OC)=CC1=O', 'COc1ccccc1CNCCCCCCNC1=CC(=O)C(NCCCCCCNCc2ccccc2OC)=CC1=O', 'COc1ccccc1CN(C)CCCCCCNC1=CC(=O)C(NCCCCCCN(C)Cc2ccccc2OC)=CC1=O', 'COc1ccccc1CN(CCCCCCNC1=CC(=O)C(NCCCCCCN(Cc2ccccc2OC)C(C)C)=CC1=O)C(C)C', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccccc2)=CC1=O)Cc1ccccc1', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccc(OC)cc2)=CC1=O)Cc1ccc(OC)cc1', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2cccc(OC)c2)=CC1=O)Cc1cccc(OC)c1', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccccc2C)=CC1=O)Cc1ccccc1C', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccccc2Cl)=CC1=O)Cc1ccccc1Cl', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccccc2[N+](=O)[O-])=CC1=O)Cc1ccccc1[N+](=O)[O-]', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccco2)=CC1=O)Cc1ccco1', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2cccs2)=CC1=O)Cc1cccs1', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccccn2)=CC1=O)Cc1ccccn1', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2cccnc2)=CC1=O)Cc1cccnc1', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)Cc2ccncc2)=CC1=O)Cc1ccncc1', 'COc1ccccc1CNCCCCCCN1CCC(CCC2CCN(CCCCCCNCc3ccccc3OC)CC2)CC1', 'COc1ccccc1CNCCCCCC(=O)N1CCC(CCC2CCN(C(=O)CCCCCNCc3ccccc3OC)CC2)CC1', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(N)c1ccc(=O)[nH]c1C2', 'CC1=C[C@H]2Cc3[nH]c(=O)ccc3[C@]3(C1)NCCC[C@H]23', 'CC1=C[C@H]2Cc3[nH]c(=O)ccc3[C@]3(C1)NCCC[C@H]23', 'O=C(CCCNc1c2c(nc3ccccc13)CCCC2)CCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(NCCNc1c2c(nc3ccccc13)CCCC2)C(=O)NCCNc1c2c(nc3ccccc13)CCCC2', 'c1ccc2c(NCCOCCOCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'O=C(c1ccccc1)N1c2ccccc2Sc2ccccc21', 'COc1ccc(C(=O)N2c3ccccc3Sc3ccccc32)cc1', 'O=C(Cc1ccccc1)N1c2ccccc2Sc2ccccc21', 'O=C(CCc1ccccc1)N1c2ccccc2Sc2ccccc21', 'O=C(c1c2ccccc2cc2ccccc12)N1c2ccccc2Sc2ccccc21', 'O=C(c1c2ccccc2cc2ccccc12)N1c2ccccc2Sc2ccccc21', 'CC(C)C1=N[C@H]2CC=C3CC4C(=O)C[C@]5(C)[C@@H]([C@H](C)N(C)C)[C@H](O)C[C@@]5(C)[C@@H]4CC[C@H]3[C@]2(C)CO1', 'CC1=N[C@H]2CC=C3C[C@H]4C(=O)C[C@]5(C)[C@@H]([C@H](C)N(C)C)[C@H](O)C[C@@]5(C)[C@@H]4CC[C@H]3[C@]2(C)CO1', 'C[C@@H]([C@H]1[C@H](O)C[C@@]2(C)[C@@H]3CC[C@@H]4C(=CC[C@@H]5N=C(C(C)(C)C)OC[C@]54C)C[C@H]3C(=O)C[C@]12C)N(C)C', 'CC[C@H](C)C1=N[C@H]2CC=C3C[C@H]4C(=O)C[C@]5(C)[C@@H]([C@H](C)N(C)C)[C@H](O)C[C@@]5(C)[C@@H]4CC[C@H]3[C@]2(C)CO1', 'CC(C)C1=N[C@H]2CC=C3C[C@H]4C(=O)C[C@]5(C)[C@@H]([C@H](C)N(C)C)[C@H](OC(=O)C(C)(C)C)C[C@@]5(C)[C@@H]4CC[C@H]3[C@]2(C)CO1', 'CN[C@@H](C)[C@H]1[C@H](O)C[C@@]2(C)[C@@H]3CC[C@@H]4C(=CC[C@@H]5N=C(C(C)C)OC[C@]54C)C[C@H]3C(=O)C[C@]12C', 'CC(C)C1=N[C@H]2CC[C@]34C[C@]35C(=O)C[C@]3(C)[C@@H]([C@H](C)N(C)C)[C@H](O)C[C@@]3(C)[C@@H]5CC[C@H]4[C@]2(C)CN1', 'CC[C@H](C)C1=N[C@H]2CC[C@]34C[C@]35C(=O)C[C@]3(C)[C@@H]([C@H](C)N(C)C)[C@H](O)C[C@@]3(C)[C@@H]5CC[C@H]4[C@]2(C)CN1', 'CC[C@H](C)C1=N[C@H]2CC=C3C[C@H]4C(=O)C[C@]5(C)[C@@H]([C@H](C)N(C)C)[C@H](O)C[C@@]5(C)[C@@H]4CC[C@H]3[C@]2(C)CN1', 'NC(=O)c1cc[n+](CCC[n+]2ccc(/C=N/O)cc2)cc1.[Br-].[Br-]', 'NC(=O)c1cc[n+](CCC[n+]2ccc(/C=N/O)cc2)cc1.[Br-].[Br-]', 'NC(=O)c1cc[n+](CCC[n+]2ccc(/C=N/O)cc2)cc1.[Br-].[Br-]', 'NC(=O)c1cc[n+](C/C=C/C[n+]2ccc(/C=N/O)cc2)cc1.[Br-].[Br-]', 'NC(=O)c1cc[n+](C/C=C/C[n+]2ccc(/C=N/O)cc2)cc1.[Br-].[Br-]', 'O=C1C(=O)c2ccc([N+](=O)[O-])c3c([N+](=O)[O-])ccc1c23', 'CCCCC1=NC2(CCCC2)C(=O)N1Cc1cc(OC)c(OC)cc1Br', 'CCCCC1=NC2(CCCC2)C(=O)N1Cc1ccc([N+](=O)[O-])cc1', 'CCCCC1=NC2(CCCC2)C(=O)N1CCN1C(=O)c2ccccc2C1=O', 'CCCCC1=NC2(CCCC2)C(=O)N1Cc1cc2c(cc1C)OCO2', 'CCCCC1=NC2(CCCC2)C(=O)N1CC(=O)c1ccc(Cl)cc1', 'CCCCC1=NC2(CCCC2)C(=O)N1Cc1ccc(-c2ccccc2-n2nnnc2C(c2ccccc2)(c2ccccc2)c2ccccc2)cc1', 'CCCCC1=NC2(CCCC2)C(=O)N1Cc1ccc(Br)cc1', 'C/C(=C\\CC[C@@H](C)[C@H]1CC[C@@]2(C)C3=CC[C@H]4C(C)(C)[C@@H](O)CC[C@]4(C)C3=CC[C@]12C)CO', 'CC(C)=CCc1c(O)cc(O)c2c1O[C@H](c1ccccc1)CC2=O', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2C1c1ccc(C)cc1)CCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2C1c1ccc(OC)cc1)CCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2C1c1ccc(OC)cc1)CCCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1ccccc1)CCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1ccccc1)CCCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1ccccc1)CCCCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1ccc(OC)cc1)CCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1ccc(OC)cc1)CCCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1ccc(OC)cc1)CCCCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1cccc(OC)c1)CCCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1ccc(OC)c(OC)c1)CCCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1ccc(F)cc1)CCCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1ccc(F)cc1)CCCCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1ccc(Cl)cc1)CCCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1cccc([N+](=O)[O-])c1)CCCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1ccc(OC)c(OC)c1)CCCCC3', 'Cc1nc2nc3c(c(N)c2c(-c2ccccc2)c1C(=O)OC(C)C)CCCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1ccncc1)CCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1ccncc1)CCCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1ccncc1)CCCCC3', 'Oc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCNC2', 'COc1cc2c(cc1OC)[C@]1(OC2=O)[C@H](O)[C@@H](O)C=C2CCN(C)[C@H]21', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31[C@H](O)CN(C)C2', 'COc1ccc2c3c1O[C@H]1C[C@H](O)C=C[C@@]31[C@H](O)CN(C)C2', 'CN1CC[C@@]23C=C[C@H](O)C[C@@H]2Oc2c(O)ccc(c23)C1', 'COc1cc2c(cc1OC)-c1cccc3c1N(CC3)C2', 'CC(=O)O[C@H]1[C@H]2c3cc4c(cc3CN3CCC(=C[C@@H]1O)[C@H]23)OCO4', 'CC(=O)O[C@H]1[C@H]2c3cc4c(cc3CN3CCC(=C[C@@H]1O)[C@H]23)OCO4', 'C[N+](C)(C)CCOC(=O)Nc1cc(Cl)nc(Cl)c1.[I-]', 'CCN(CCCCCC(=O)N1CCC(C2CCN(C(=O)CCCCCN(CC)Cc3ccccc3OC)CC2)CC1)Cc1ccccc1OC', 'CCN(CCCCCC(=O)N1CCC(C2CCN(C(=O)CCCCCN(CC)Cc3ccccc3OC)CC2)CC1)Cc1ccccc1OC', 'CCN(CCCCCC(=O)Nc1ccc(-c2ccc(NC(=O)CCCCCN(CC)Cc3ccccc3OC)cc2)cc1)Cc1ccccc1OC', 'COc1ccccc1CNCCCNCCCNCCC[n+]1c(-c2ccccc2)c2cc(N)ccc2c2ccc(N)cc21', 'C#CCN[C@@H]1CCc2cc(Cl)c(OC(=O)N(C)CCC)cc21', 'C#CCN[C@@H]1CCc2ccc(OC(=O)N(C)CC)cc21', 'C#CCN[C@@H]1CCc2ccc(OC(=O)N(C)CC)cc21', 'C#CCN[C@@H]1CCc2ccc(OC(=O)N(C)CC)cc21', 'NCCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'CC[C@@]1(c2cccc(O)c2)CCCCN(CCCCCCCCCCN2CCCC[C@@](CC)(c3cccc(O)c3)C2)C1', 'CC[C@@]1(c2cccc(O)c2)CCCCN(CCCCCCCCCN2CCCC[C@@](CC)(c3cccc(O)c3)C2)C1', 'COc1ccc2c(c[n+](C)c3c4cc5c(cc4ccc23)OCO5)c1OC.[Cl-]', 'COc1ccc2c(c[n+](C)c3c4cc5c(cc4ccc23)OCO5)c1OC.[Cl-]', 'COc1ccc2c(c[n+](C)c3c4cc5c(cc4ccc23)OCO5)c1OC.[Cl-]', 'Cc1ccc2[nH]c3ccncc3c2c1', 'OCCC1CCN(Cc2ccc(OCCCN3CCCCC3)cc2)CC1', 'OCC1CCN(Cc2ccc(OCCCN3CCCCC3)cc2)CC1', 'CCN(Cc1ccccc1)Cc1ccc(-c2cc3cc(OCCNC(=O)Cc4cc(F)cc(F)c4)ccc3oc2=O)cc1', 'CCN(Cc1ccccc1)Cc1ccc(-c2cc3cc(OCCNC(=O)/C=C/c4cc(F)cc(F)c4)ccc3oc2=O)cc1', 'CCN(Cc1ccccc1)Cc1ccc(-c2cc3cc(OCCNC(=O)/C=C/c4cc(F)cc(F)c4)ccc3oc2=O)cc1', 'CCN(Cc1ccccc1)Cc1ccc(-c2cc3cc(OCCNC(=O)CCc4ccc(F)c(F)c4)ccc3oc2=O)cc1', 'CCN(Cc1ccccc1)Cc1ccc(-c2cc3ccc(OCCNC(=O)/C=C/c4cc(F)cc(F)c4)cc3oc2=O)cc1', 'CCN(Cc1ccccc1)Cc1ccc(-c2cc3ccc(OCCNC(=O)CCc4ccc(F)c(F)c4)cc3oc2=O)cc1', 'CCN(Cc1ccccc1)Cc1ccc(-c2cc3ccc(OCCNC(=O)/C=C/c4ccc(Cl)c(Cl)c4)cc3oc2=O)cc1', 'CCN(Cc1ccccc1)Cc1ccc(-c2cc3ccc(OCCNC(=O)Cc4cc(F)cc(F)c4)cc3oc2=O)cc1', 'O=C(CO/N=[N+](\\[O-])N1CCCC1)NCCCNc1c2c(nc3ccccc13)CCCC2', 'O=[N+]([O-])OCCCCCCNCCCCNc1c2c(nc3ccccc13)CCCC2', 'O=[N+]([O-])OCCCNCCCCNc1c2c(nc3ccccc13)CCCC2', 'O=[N+]([O-])OCCNCCCCNc1c2c(nc3ccccc13)CCCC2', 'O=[N+]([O-])OCCCCCCNCCCNc1c2c(nc3ccccc13)CCCC2', 'O=[N+]([O-])OCCCNCCCNc1c2c(nc3ccccc13)CCCC2', 'O=[N+]([O-])OCCNCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(CCO[N+](=O)[O-])NCCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(CO[N+](=O)[O-])NCCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(CCO[N+](=O)[O-])NCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(CO[N+](=O)[O-])NCCCNc1c2c(nc3ccccc13)CCCC2', 'CC(C)(C)C(=O)NCCCCNc1c2c(nc3ccccc13)CCCC2', 'CC(C)(CO[N+](=O)[O-])C(=O)NCCCCNc1c2c(nc3ccccc13)CCCC2', 'CC(C)(CO[N+](=O)[O-])C(=O)NCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(CO/N=[N+](\\[O-])N1CCCC1)NCCCCNc1c2c(nc3ccccc13)CCCC2', 'COc1cc2c3cc1Oc1c(O)c(OC)cc4c1[C@@H](Cc1ccc(O)c(c1)Oc1ccc(cc1)C[C@@H]3N(C)CC2)N(C)CC4', 'Br.COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'Br.COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'Br.COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(C)C2', 'COc1ccc2c(=O)oc3c(OC)cc(CCN(C)C)c4c(=O)oc1c2c34', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)CC[C@@]31CCN(C)C2', 'CC(C)=CCc1c(-c2ccc(O)cc2O)oc2c3c(cc(O)c2c1=O)OC(C)(C)C=C3', 'CC(C)=CCc1c(-c2ccc(O)cc2O)oc2c(CC=C(C)C)c(O)cc(O)c2c1=O', 'COc1cc([C@H]2OC[C@H]3[C@@H]2CO[C@@H]3c2cc(OC)c(OC)c(OC)c2)cc(OC)c1OC', 'CC(C)=CCOc1c2ccoc2cc2oc(=O)ccc12', 'CC(C)(O[C@@H]1O[C@H](CO)[C@@H](O)[C@H](O)[C@H]1O)[C@H]1Cc2cc3ccc(=O)oc3cc2O1', 'C=CC(=O)N1C/C(=C\\c2ccc(C)cc2)C(=O)/C(=C/c2ccc(C)cc2)C1', 'C=CC(=O)N1C/C(=C\\c2ccc(C)cc2)C(=O)/C(=C/c2ccc(C)cc2)C1', 'O=c1c2ccccc2nc2n1CCC2', 'O=c1c2ccccc2nc2n1CCC2', 'O=c1c2ccccc2nc2n1CCC2', 'CN(C)C(=O)Oc1cc(OC(=O)N(C)C)cc(C(O)CNC(C)(C)C)c1', 'CN(C)C(=O)Oc1cc(OC(=O)N(C)C)cc(C(O)CNC(C)(C)C)c1', 'CN(C)C(=O)Oc1cc(OC(=O)N(C)C)cc(C(O)CNC(C)(C)C)c1', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1cccc(CN(C)C)c1)C2', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(CN(C)C)cc1)C2', 'CCN(C)Cc1cccc(/C=C2\\Cc3cc(OC)c(OC)cc3C2=O)c1', 'CCN(C)Cc1ccc(/C=C2\\Cc3cc(OC)c(OC)cc3C2=O)cc1', 'CCN(CC)Cc1cccc(/C=C2\\Cc3cc(OC)c(OC)cc3C2=O)c1', 'CCN(CC)Cc1ccc(/C=C2\\Cc3cc(OC)c(OC)cc3C2=O)cc1', 'CCN(CC)Cc1ccc(/C=C2\\Cc3cc(OC)c(OC)cc3C2=O)cc1', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1cccc(CN3CCCC3)c1)C2', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(CN3CCCC3)cc1)C2', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1cccc(CN3CCCCC3)c1)C2', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(CN3CCCCC3)cc1)C2', 'COc1cc2c(cc1OC)C(=O)C(Cc1cccc(CN(C)C)c1)C2', 'COc1cc2c(cc1OC)C(=O)C(Cc1ccc(CN(C)C)cc1)C2', 'CCN(C)Cc1cccc(CC2Cc3cc(OC)c(OC)cc3C2=O)c1', 'CCN(C)Cc1ccc(CC2Cc3cc(OC)c(OC)cc3C2=O)cc1', 'CCN(CC)Cc1cccc(CC2Cc3cc(OC)c(OC)cc3C2=O)c1', 'CCN(CC)Cc1ccc(CC2Cc3cc(OC)c(OC)cc3C2=O)cc1', 'COc1cc2c(cc1OC)C(=O)C(Cc1cccc(CN3CCCC3)c1)C2', 'COc1cc2c(cc1OC)C(=O)C(Cc1ccc(CN3CCCC3)cc1)C2', 'COc1cc2c(cc1OC)C(=O)C(Cc1cccc(CN3CCCCC3)c1)C2', 'COc1cc2c(cc1OC)C(=O)C(Cc1ccc(CN3CCCCC3)cc1)C2', 'COc1ccc(-c2csc(NC(=O)CCN3CCCC3)n2)cc1', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2[C@@H]1c1ccc(OC)cc1)CCCC3', 'COc1ccccc1CN(C)Cc1ccc(C(=O)c2ccc(OC)c(OC)c2)cc1', 'COc1ccc(C(=O)c2ccc(CN(C)Cc3cccc(C)c3)cc2)cc1OC', 'CCN(Cc1ccccc1)Cc1ccc(C(=O)c2ccc(OC)c(OC)c2)cc1', 'COc1ccc(C(=O)c2ccc(CN(CCO)Cc3ccccc3)cc2)cc1OC', 'COc1ccc(C(=O)c2ccc(/C=C/CN(C)Cc3ccccc3)cc2)cc1OC', 'COc1ccc(C(=O)c2ccc(CN(C)Cc3ccccc3)cc2)cc1', 'COc1ccc(C(=O)c2ccc(CN(C)Cc3ccccc3)cc2)cc1', 'COc1ccc(C(=O)c2ccc(CN(C)Cc3ccccc3)cc2)cc1F', 'CN(Cc1ccccc1)Cc1ccc(C(=O)c2ccc3ccccc3c2)cc1', 'COc1ccc(C(=O)c2ccc(CN(CCN(C)C)Cc3ccccc3)cc2)cc1OC', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccccc1C(F)(F)F)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccc(-c2ccccc2)cc1)CCCC3', 'COc1ccc(C(=O)c2ccc(CN(C)Cc3ccccc3)cc2)cc1OC', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccc(OC)c(OC)c1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1cccnc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccncc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2[C@H]1c1ccc(OC)cc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2[C@H]1c1ccc(OC)cc1)CCCC3', 'CCN(CC)Cc1ccc(C(=O)c2ccc(OC)c(OC)c2)cc1', 'COc1ccc(C(=O)c2ccc(CN(C)CCN(C)C)cc2)cc1OC', 'C#CCN(C)Cc1ccc(C(=O)c2ccc(OC)c(OC)c2)cc1', 'COc1ccc(C(=O)c2ccc(CN3CCc4ccccc4C3)cc2)cc1OC', 'COc1ccc(C(=O)c2ccc(CN(C)Cc3cccc([N+](=O)[O-])c3)cc2)cc1OC', 'C=CC[N+](C)(C)c1ccc(CCC(=O)CCc2ccc([N+](C)(C)CC=C)cc2)cc1.[Br-].[Br-]', 'C=CC[N+](C)(C)c1ccc(CCC(=O)CCc2ccc([N+](C)(C)CC=C)cc2)cc1.[Br-].[Br-]', 'C=CC[N+](C)(C)c1ccc(CCC(=O)CCc2ccc([N+](C)(C)CC=C)cc2)cc1.[Br-].[Br-]', 'C=CC[N+](C)(C)c1ccc(CCC(=O)CCc2ccc([N+](C)(C)CC=C)cc2)cc1.[Br-].[Br-]', 'C=CC[N+](C)(C)c1ccc(CCC(=O)CCc2ccc([N+](C)(C)CC=C)cc2)cc1.[Br-].[Br-]', 'C=CC[N+](C)(C)c1ccc(CCC(=O)CCc2ccc([N+](C)(C)CC=C)cc2)cc1.[Br-].[Br-]', 'C=CC[N+](C)(C)c1ccc(CCC(=O)CCc2ccc([N+](C)(C)CC=C)cc2)cc1.[Br-].[Br-]', 'C=CC[N+](C)(C)c1ccc(CCC(=O)CCc2ccc([N+](C)(C)CC=C)cc2)cc1.[Br-].[Br-]', 'C=CC[N+](C)(C)c1ccc(CCC(=O)CCc2ccc([N+](C)(C)CC=C)cc2)cc1.[Br-].[Br-]', 'C=CC[N+](C)(C)c1ccc(CCC(=O)CCc2ccc([N+](C)(C)CC=C)cc2)cc1.[Br-].[Br-]', 'C=CC[N+](C)(C)c1ccc(CCC(=O)CCc2ccc([N+](C)(C)CC=C)cc2)cc1.[Br-].[Br-]', 'C=CC[N+](C)(C)c1ccc(CCC(=O)CCc2ccc([N+](C)(C)CC=C)cc2)cc1.[Br-].[Br-]', 'C=CC[N+](C)(C)c1ccc(CCC(=O)CCc2ccc([N+](C)(C)CC=C)cc2)cc1.[Br-].[Br-]', 'COC(=O)CC[C@@H](C)[C@H]1CC(=O)[C@@]2(C)C3=C(C(=O)[C@@H](OC(C)=O)[C@]12C)[C@@]1(C)CC[C@H](O)C(C)(C)[C@@H]1CC3=O', 'C[C@H](CCC(=O)O)[C@H]1CC(=O)[C@@]2(C)C3=C(C(=O)C[C@]12C)[C@@]1(C)CCC(=O)C(C)(C)[C@@H]1C[C@@H]3O', 'CC(C)(O)[C@@H]1Cc2cc3ccc(=O)oc3cc2O1', 'C[C@H](CCC(=O)O)[C@H]1CC(=O)[C@@]2(C)C3=C(C(=O)C[C@]12C)[C@@]1(C)CC[C@H](O)C(C)(C)[C@@H]1C[C@@H]3O', 'C[C@H](CC(=O)C[C@@H](C)[C@H]1CC(=O)[C@@]2(C)C3=C(C(=O)C[C@]12C)[C@@]1(C)CCC(=O)C(C)(C)[C@@H]1CC3=O)C(=O)O', 'COc1cc([C@H]2Oc3cc(O)c(C/C=C(\\C)CCC=C(C)C)c(O)c3C(=O)[C@@H]2O)ccc1O', 'COc1cc([C@H]2Oc3cc(O)c(C/C=C(\\C)CCC=C(C)C)c(O)c3C(=O)[C@@H]2O)ccc1O', 'COc1ccccc1CNCCCCCCNCCCCCCCCNCCCCCCNCc1ccccc1OC.Cl.Cl.Cl.Cl', 'COc1ccc([C@H]2OC[C@H]3[C@@H]2CO[C@@H]3c2cc(OC)c(OC)c(OC)c2)cc1OC', 'Cc1ccc2c(c1)C(=O)C(=O)N2Cc1ccc(Br)cc1', '[O-]c1cc2c3c(c1)c1cc4c(cc1c[n+]3CC2)OCO4', 'CC1(C)Oc2cc3oc(=O)ccc3cc2C[C@@H]1O', 'C=C[C@@H]1[C@H]2C=C(C)C[C@]1(NC)c1ccc(=O)[nH]c1C2', 'O=C(CCCCCCCCc1ccc(O)cc1)c1c(O)cccc1O', 'O=C(CCCCCCCCc1ccc(O)cc1)c1c(O)cccc1O', 'O=C(CCCCCCCCc1ccc(O)c(O)c1)c1c(O)cccc1O', 'O=C(CCCCCCCCc1ccc(O)c(O)c1)c1c(O)cccc1O', 'O=C(OC[C@H]1O[C@@H](OC(=O)c2cc(O)c(O)c(O)c2)[C@H](OC(=O)c2cc(O)c(O)c(O)c2)[C@@H](OC(=O)c2cc(O)c(O)c(O)c2)[C@@H]1O)c1cc(O)c(O)c(O)c1', 'CN(CCCCCCCOc1ccc(-c2cc3ccccc3o2)cc1)Cc1ccccc1', 'CC(=O)c1c(-c2ccc(OCCCCCCCN(C)Cc3ccccc3)cc2)oc2ccccc12', 'CN(CCCCCCCOc1ccc(-c2oc3ccccc3c2C(=O)c2ccccc2)cc1)Cc1ccccc1', 'Cc1ccc(C(=O)c2c(-c3ccc(OCCCCCCCN(C)Cc4ccccc4)cc3)oc3ccccc23)cc1', 'Clc1ccc2c(NCCCNCC34CC5CC(CC(C5)C3)C4)ccnc2c1', 'CC1(C)CC(=O)C2=C(C1)Nc1nc3c(c(N)c1C2c1ccccc1)CCCC3', 'CC1(C)CC(=O)C2=C(C1)Nc1nc3c(c(N)c1C2c1ccc(F)cc1)CCCC3', 'Cc1ccc(C2C3=C(CC(C)(C)CC3=O)Nc3nc4c(c(N)c32)CCCC4)cc1', 'COc1ccc(C2C3=C(CC(C)(C)CC3=O)Nc3nc4c(c(N)c32)CCCC4)cc1', 'CC1(C)CC(=O)C2=C(C1)Nc1nc3c(c(N)c1C2c1ccncc1)CCCC3', 'Cc1ccc(Nc2nccc(N(C)c3ccc4c(C)n(C)nc4c3)n2)cc1S(N)(=O)=O', 'Clc1ccc2c(NCCCNCCCOc3cccc4[nH]c5ccccc5c34)c3c(nc2c1)CCCC3', 'Clc1ccc2c(NCCCNCCCCOc3cccc4[nH]c5ccccc5c34)c3c(nc2c1)CCCC3', 'Clc1ccc2c(NCCCNCCCCCOc3cccc4[nH]c5ccccc5c34)c3c(nc2c1)CCCC3', 'Clc1ccc2c(NCCCNCCCCCCOc3cccc4[nH]c5ccccc5c34)c3c(nc2c1)CCCC3', 'CN[C@@H](Cc1ccccc1)C(=O)NCC(=O)Nc1c2c(nc3ccccc13)CCCC2', 'N[C@@H](Cc1ccc2ccccc2c1)C(=O)NCC(=O)Nc1c2c(nc3ccccc13)CCCC2', 'O=C(CNC(=O)[C@@H]1Cc2ccccc2CN1)Nc1c2c(nc3ccccc13)CCCC2', 'N[C@@H](Cc1ccccc1)C(=O)NCC(=O)Nc1c2c(nc3ccccc13)CCCC2', 'N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCC(=O)Nc1c2c(nc3ccccc13)CCCC2', 'CN[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCC(=O)Nc1c2c(nc3ccccc13)CCCC2', 'CC(C)(C)OC(=O)N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCC(=O)Nc1c2c(nc3ccccc13)CCCC2', 'N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCC(=O)Nc1c2c(nc3ccccc13)CCCC2', 'CC(C)(C)OC(=O)N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCNc1c2c(nc3ccccc13)CCCC2', 'N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCNc1c2c(nc3ccccc13)CCCC2', 'CC(C)(C)OC(=O)N[C@@H](CCC(=O)NCCNc1c2c(nc3ccccc13)CCCC2)C(=O)NCCNc1c2c(nc3ccccc13)CCCC2', 'N[C@@H](CCC(=O)NCCNc1c2c(nc3ccccc13)CCCC2)C(=O)NCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(NCCCCCNc1c2c(nc3ccccc13)CCCC2)[C@@H]1CCCN1CCNc1c2c(nc3ccccc13)CCCC2', 'CO[C@@H]1[C@@H](O)[C@H]2c3cc4c(cc3CN3CC=C([C@H]1O)[C@H]23)OCO4', 'O=C(O)C(F)(F)F.O=C1C=C2S[C@H]3C[C@]2(C=C1Br)C1=C(N3)C(=O)c2[nH]cc3c2C1=NCC3', 'O=C(O)C(F)(F)F.O=C1C=C2S[C@H]3C[C@]2(C=C1Br)C1=C(N3)C(=O)c2[nH]cc3c2C1=NCC3', 'Cl.Nc1c2c(nc3ccccc13)CCCC2', 'Cl.Nc1c2c(nc3ccccc13)CCCC2', 'Cl.Nc1c2c(nc3ccccc13)CCCC2', 'Cl.Nc1c2c(nc3ccccc13)CCCC2', 'Cl.Nc1c2c(nc3ccccc13)CCCC2', 'Cl.Nc1c2c(nc3ccccc13)CCCC2', 'Cl.Nc1c2c(nc3ccccc13)CCCC2', 'Cl.Nc1c2c(nc3ccccc13)CCCC2', 'Cl.Nc1c2c(nc3ccccc13)CCCC2', 'Cl.Nc1c2c(nc3ccccc13)CCCC2', 'Cl.Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Cl.Nc1c2c(nc3cc(Cl)ccc13)CCCC2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2.Cl', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2.Cl', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2.Cl', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2.Cl', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2.Cl', 'CC(=O)N(CCCCNc1c2c(nc3ccccc13)CCCC2)CCCSc1c2c(nc3ccccc13)CCCC2', 'CC(=O)N(CCCCNc1c2c(nc3ccccc13)CCCC2)CCCSc1c2c(nc3ccccc13)CCCC2', 'c1ccc2c(NCCCCNCCCSc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'CC(=O)N(CCCCNc1c2c(nc3ccccc13)CCCC2)CCCNc1c2c(nc3ccccc13)CCCC2', 'CN(CCCCNc1c2c(nc3ccccc13)CCCC2)CCCNc1c2c(nc3ccccc13)CCCC2', 'CN(CCCCNc1c2c(nc3ccccc13)CCCC2)CCCNc1c2c(nc3ccccc13)CCCC2', 'CCN(CCCCNc1c2c(nc3ccccc13)CCCC2)CCCNc1c2c(nc3ccccc13)CCCC2', 'C=CCN(CCCCNc1c2c(nc3ccccc13)CCCC2)CCCNc1c2c(nc3ccccc13)CCCC2', 'OCCN(CCCCNc1c2c(nc3ccccc13)CCCC2)CCCNc1c2c(nc3ccccc13)CCCC2', 'c1ccc2c(NCCCN3CCN(CCCNc4c5c(nc6ccccc46)CCCC5)CC3)c3c(nc2c1)CCCC3', 'c1cc(CCNc2c3c(nc4ccccc24)CCCC3)cc(CCNc2c3c(nc4ccccc24)CCCC3)c1', 'c1cc(CCNc2c3c(nc4ccccc24)CCCC3)nc(CCNc2c3c(nc4ccccc24)CCCC3)c1', 'c1ccc2c(NCCc3ccc(CCNc4c5c(nc6ccccc46)CCCC5)cc3)c3c(nc2c1)CCCC3', 'Cc1cc(CCNc2c3c(nc4ccccc24)CCCC3)cc(CCNc2c3c(nc4ccccc24)CCCC3)c1', 'c1ccc2c(NCCc3cocc3CCNc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'CCOP(=O)(C#N)N(C)C', 'CCc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CC[N+](C)([O-])[C@@H]1N2C', 'CC(C)=CC(=O)N[C@H]1CC[C@@]2(C)[C@@H](CC[C@@H]3[C@@H]2CC[C@]2(C)C([C@H](C)N(C)C)=CC[C@@H]32)[C@H]1O', 'C[C@@H]([C@H]1CC=C2[C@@H]3CCC4=C[C@@H](N(C)C)CC[C@]4(C)[C@H]3CC[C@@]21C)N(C)C', 'C[C@@H]([C@H]1CC[C@H]2[C@@H]3CC[C@H]4C[C@@H](N)CC[C@]4(C)[C@H]3CC[C@]12C)N(C)C', 'C[C@@H]([C@H]1CC[C@H]2[C@@H]3CC[C@H]4C[C@@H](N(C)C)CC[C@]4(C)[C@H]3CC[C@]12C)N(C)C', 'CN[C@@H](C)[C@H]1CC[C@H]2[C@@H]3CC=C4C(=O)[C@@H](NC(=O)C=C(C)C)CC[C@]4(C)[C@H]3CC[C@]12C', 'C/C=C(\\C)C(=O)N[C@H]1CC[C@]2(C)[C@H]3CC[C@]4(C)[C@@H]([C@H](C)NC)CC[C@H]4[C@@H]3CC[C@H]2[C@H]1OC(C)=O', 'Oc1cccc(C2=Nc3ccccc3SC(c3ccc(Cl)cc3)C2)c1', 'Oc1cccc(C2=Nc3ccccc3SC(c3ccccc3F)C2)c1', 'Oc1cccc(C2=Nc3ccccc3SC(c3cccc(O)c3)C2)c1', 'Cc1ccc(C2CC(c3cccc(O)c3)=Nc3ccccc3S2)cc1', 'Oc1cccc(C2=Nc3ccccc3SC(c3cccs3)C2)c1', 'Oc1cccc(C2=Nc3ccccc3SC(c3ccccc3)C2)c1', 'CCN(CCCCCC(=O)N1CCN(C(=O)CCCCCN(CC)Cc2ccccc2OC)CC1)Cc1ccccc1OC', 'CCN(CCCCCC(=O)N[C@H]1CCCC[C@H]1NC(=O)CCCCCN(CC)Cc1ccccc1OC)Cc1ccccc1OC', 'CCN(CCCCCC(=O)N[C@H]1CCCC[C@@H]1NC(=O)CCCCCN(CC)Cc1ccccc1OC)Cc1ccccc1OC', 'CCN(CCCCCC(=O)N[C@H]1CC[C@H](NC(=O)CCCCCN(CC)Cc2ccccc2OC)CC1)Cc1ccccc1OC', 'CCN(CCCCCC(=O)N1Cc2ccc3c4c(ccc(c24)C1)CN(C(=O)CCCCCN(CC)Cc1ccccc1OC)C3)Cc1ccccc1OC', 'CCN(CCCCCCN1CCN(CCCCCCN(CC)Cc2ccccc2OC)C(=O)C1=O)Cc1ccccc1OC', 'CCN(CCCCCCN1CC(=O)N(CCCCCCN(CC)Cc2ccccc2OC)CC1=O)Cc1ccccc1OC', 'CCN(CCCCCCn1c(=O)c(=O)n(CCCCCCN(CC)Cc2ccccc2OC)c2ccccc21)Cc1ccccc1OC', 'CCN(CCCCCCN1CCN(CCCCCCN(CC)Cc2ccccc2OC)C1=O)Cc1ccccc1OC', 'CCN(CCCCCCN1C(=O)c2ccc(-c3ccc4c(c3)C(=O)N(CCCCCCN(CC)Cc3ccccc3OC)C4=O)cc2C1=O)Cc1ccccc1OC', 'CCN(CCCCCCn1c(=O)c2cc3c(=O)n(CCCCCCN(CC)Cc4ccccc4OC)c(=O)c3cc2c1=O)Cc1ccccc1OC', 'CCN(CCCCCCN1C(=O)c2ccc3c4c(ccc(c24)C1=O)C(=O)N(CCCCCCN(CC)Cc1ccccc1OC)C3=O)Cc1ccccc1OC', 'CCN(CCCCCCN1C(=O)c2ccc3c4c(ccc(c24)C1=O)C(=O)N(CCCCCCN(CC)Cc1ccccc1OC)C3=O)Cc1ccccc1OC', 'CCN(CCCCCCN1C(=O)c2cccc3cccc(c23)C1=O)Cc1ccccc1OC', 'CCn1c(C)nc2cc(C(=O)CCC3CCN(Cc4ccccc4)CC3)ccc21.Cl', 'Cc1nc2cc(C(=O)CCC3CCN(Cc4ccccc4)CC3)ccc2[nH]1.Cl', 'Cc1nc2ccc(C(=O)CCC3CCN(Cc4ccccc4)CC3)cc2s1.Cl', 'Cl.O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccccc1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2cccc(Cc3ccccc3)c2)c1.Cl', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc(C(=O)c3ccccc3)cc2)c1.Cl', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1.Cl', 'CNC(=O)Oc1cccc(CN(C)CC#CCOc2ccc3c(=O)c4ccccc4oc3c2)c1.Cl', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3ccccc3c2)c1.Cl', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1.Cl', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1.Cl', 'CCC1=CC2CC(C1)Cc1nc3ccccc3c(N)c12.Cl', 'CCN(C)C(=O)Oc1cccc2c1C(N)CC2.Cl', 'C#CCNC1CCc2cccc(OC(=O)N(C)CCC)c21.Cl', 'C#CCNC1CCc2ccc(OC(=O)N(C)C3CCCCC3)cc21.Cl', 'C#CCNC1CCc2ccc(OC(=O)N(C)c3ccccc3)cc21.Cl', 'COc1ccc(N(C)C(=O)Oc2ccc3c(c2)C(N)CC3)cc1.Cl', 'C#CCNC1CCc2ccc(OC(=O)NCCC)cc21.Cl', 'CCCCCCN(C)C(=O)Oc1ccc2c(c1)C(N)CC2.Cl', 'C#CCN(C)C1CCc2cccc(OC(=O)N(C)C)c21.Cl', 'C#CCNC1CCc2ccc(OC(=O)N(C)CCCCCC)cc21.Cl', 'CCN(C)C(=O)Oc1ccc2c(c1)C(N)CC2.Cl', 'CN(C)C(=O)Oc1ccc2c(c1)C(N)CC2.Cl', 'C#CCN(CC)C1CCc2ccc(OC(=O)N(C)C)cc21.Cl', 'C#CCNC1CCc2ccc(OC(=O)N(CC)C3CCCCC3)cc21.Cl', 'C#CCNC1CCc2ccc(OC(=O)NCC)cc21.Cl', 'CCCNC(=O)Oc1ccc2c(c1)C(N)CC2.Cl', 'CN(C)C(=O)Oc1cccc2c1CCC2N.Cl', 'C#CCNC1CCc2ccc(OC(=O)N(C)C)cc21.Cl', 'C#CCNC1CCc2c(Cl)cc(OC(=O)N(C)CC)cc21.Cl', 'CN(C(=O)Oc1ccc2c(c1)C(N)CC2)C1CCCCC1.Cl', 'CCN(C)C(=O)Oc1cccc2c1CCC2N.Cl', 'CN(C)C(=O)Oc1cccc2c1CCC2N(C)C.Cl', 'C#CCNC1CCc2ccc(OC(=O)N(C)CC)cc21.Cl', 'CNC1CCc2ccc(OC(=O)N(C)C)cc21.Cl', 'C#CCNC1CCc2ccc(OC(=O)N(C)CCC)cc21.Cl', 'C#CCNC1CCc2c(OC(=O)N(C)CC)cccc21.Cl', 'CCNC1CCc2ccc(OC(=O)N(C)C)cc21.Cl', 'CCN(C)C(=O)Oc1ccc2c(c1)C(NC)CC2.Cl', 'C#CCNC1CCc2ccc(OC(=O)N(C)Cc3ccccc3)cc21.Cl', 'C#CCNC1CCc2ccc(OC(=O)N(C)c3ccc(OC)cc3)cc21.Cl', 'C#CCNC1CCc2c(OC(=O)N(C)C)cccc21.Cl', 'C#CCN(C)C1CCc2c(OC(=O)N(C)CC)cccc21.Cl', 'CCNC(=O)Oc1ccc2c(c1)C(N)CC2.Cl', 'CCCN(C)C(=O)Oc1ccc2c(c1)C(N)CC2.Cl', 'CN(C)C(=O)Oc1ccc2c(c1)C(N(C)C)CC2.Cl', 'Cl.c1ccc(-c2ccc(NCCCCCN3CCc4ccccc4C3)nn2)cc1', 'CCC1=CC2Cc3nc4ccccc4c(NCCCCCCCCNc4c5c(nc6ccccc46)CCCC5)c3C(C1)C2.Cl.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3C(C1)C2.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCNc4c5c(nc6ccccc46)CCCC5)c3C(C1)C2.Cl.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCNc4c5c(nc6ccccc46)CCCC5)c3C(C1)C2.Cl.Cl', 'COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(C)Cc1ccccc1OC.Cl.Cl', 'Cl.O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2[nH]c(-c3ccccc3)nc2c1', 'Cl.O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)OCCO2', 'Cc1c(C(=O)NN2CCCCC2)nn(-c2ccc(Cl)cc2Cl)c1-c1ccc(Cl)cc1.Cl', 'CNC(=O)Oc1ccc2c(c1)CN(CCCOc1ccc3c(=O)c4ccccc4oc3c1)CC2.Cl', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1.Cl', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccccc2)c1.Cl', 'CCC1=CC2Cc3nc4ccccc4c(N)c3C(C1)C2.Cl', 'CC1=CC2Cc3nc4ccccc4c(N)c3C(C1)C2.Cl', 'C#CCNC1CCc2cccc(OC(=O)N(C)C)c21.Cl', 'C#CCN(C)C1CCc2ccc(OC(=O)N(C)C)cc21.Cl', 'C#CCNC1CCc2ccc(OC(=O)N(CC)CCCC)cc21.Cl', 'CN(CCCCCNc1ccc(-c2ccccc2)nn1)Cc1ccccc1.Cl', 'C#CCNC1CCc2ccc(OC(=O)N(C)CCCC)cc21.Cl', 'CN(C)C(=O)Oc1cccc2c1C(N)CC2.Cl', 'CC1=CC2CC(C1)Cc1nc3ccccc3c(N)c12.Cl', 'CNC1CCc2cccc(OC(=O)N(C)C)c21.Cl', 'CCCNC1CCc2ccc(OC(=O)N(C)CC)cc21.Cl', 'C#CCNC1CCc2c(Cl)cc(OC(=O)N(C)CCC)cc21.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCNc4c5c(nc6ccccc46)CCCC5)c3C(C1)C2.Cl.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)c3C(C1)C2.Cl.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)c3C(C1)C2.Cl.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)c3C(C1)C2.Cl.Cl', 'C[n+]1ccn(CCN2CCCC2)c1/C=N/O.[Cl-]', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CC[N@@+](C)([O-])[C@@H]1N2C.Cl', 'CN1c2ccc(OC(=O)Nc3ccccc3)cc2[C@]2(C)CC[N@@+](C)([O-])[C@H]12.Cl', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CC[N@@+](C)([O-])[C@@H]1N2C.Cl', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CC[N+](C)(C)C1N2C.COS(=O)(=O)[O-]', 'CCc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CC[N+](C)([O-])[C@@H]1N2C.Cl', 'CC[N+](CC)(CC)CCOc1ccc(C(=O)NCCCCCC(=O)NNc2c3c(nc4cc(Cl)ccc24)CCCC3)cc1.[Br-]', 'CC[N+](CC)(CC)CCOc1cc(OCC[N+](CC)(CC)CC)cc(C(=O)NCCCCCC(=O)NNc2c3c(nc4cc(Cl)ccc24)CCCC3)c1.[Br-].[Br-]', 'CC[N+](CC)(CC)CCOc1cc(OCC[N+](CC)(CC)CC)c(C(=O)NCCCCCC(=O)NNc2c3c(nc4cc(Cl)ccc24)CCCC3)cc1OCC[N+](CC)(CC)CC.[Br-].[Br-].[Br-]', 'CCc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CC[N@@+](C)([O-])[C@@H]1N2C.Cl', 'Cn1cc[n+](COC[n+]2ccccc2/C=N/O)c1/C=N/O.[Cl-].[Cl-]', 'C=C1C[C@]23C[C@H]1CC[C@H]2[C@]1(C)CCC[C@](C)(CO)[C@@H]1C[C@@H]3O', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)[C@@H](C)c2ccccc2OC)=CC1=O)[C@H](C)c1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)[C@@H](C)c2ccccc2OC)=CC1=O)[C@H](C)c1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)[C@H](C)c2ccccc2OC)=CC1=O)[C@H](C)c1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)[C@H](C)c2ccccc2OC)=CC1=O)[C@H](C)c1ccccc1OC', 'C[C@]1(CO)CCC[C@]2(C)[C@H]1C[C@H](O)[C@@]13C[C@@H](CC[C@@H]21)[C@]1(C)O[C@@H]13', 'CC(=O)O[C@H]1[C@H]2c3cc4c(cc3CN3CCC(=C[C@@H]1O[Si](C)(C)C(C)(C)C)[C@H]23)OCO4', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)[C@@H](C)c2ccccc2OC)=CC1=O)[C@@H](C)c1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCN(CC)[C@@H](C)c2ccccc2OC)=CC1=O)[C@@H](C)c1ccccc1OC', 'CC(C)(C)[Si](C)(C)O[C@H]1C=C2CCN3Cc4cc5c(cc4[C@H]([C@@H]1O)[C@@H]23)OCO5', 'CC1=CC2Cc3nc4oc(C)c(C)c4c(N)c3C(C1)C2', 'CCC1=CC2Cc3nc4oc(C)c(C)c4c(N)c3C(C1)C2', 'CC1=CC2Cc3nc4oc(C5CC5)nc4c(N)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCN=[N+]=[N-])c3C(C1)C2', 'CCC1=CC2Cc3nc4cc(Cl)ccc4c(NCCN=[N+]=[N-])c3C(C1)C2', 'C#CCCCCCCN1CCc2cc(OC)c(OC)cc2C1c1cccc([N+](=O)[O-])c1', 'CCC1=CC2Cc3nc4oc(C5CC5)nc4c(N)c3C(C1)C2', 'CC1=CC2Cc3nc4oc(C5CCC5)nc4c(N)c3C(C1)C2', 'COc1cc2c(cc1OC)C(c1cccc([N+](=O)[O-])c1)N(CCCCCCc1cn(CCNc3c4c(nc5cc(Cl)ccc35)CC3C=C(C)CC4C3)nn1)CC2', 'CCC1=CC2Cc3nc4cc(Cl)ccc4c(NCCn4cc(CCCCCCN5CCc6cc(OC)c(OC)cc6C5c5cccc([N+](=O)[O-])c5)nn4)c3C(C1)C2', 'CCCSc1nc2nc3c(c(N)c2s1)C1CC(C)=CC(C3)C1', 'CCCSc1nc2nc3c(c(N)c2s1)C1CC(CC)=CC(C3)C1', 'CSc1n[nH]c2nc3c(c(N)c12)C1CC(C)=CC(C3)C1', 'CC1=CC2Cc3nc4[nH]ncc4c(N)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NC(=O)CCl)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(O)c3C(C1)C2', 'C#CCCCCCCN1CCc2cc(OC)c(OC)cc2C1c1ccc([N+](=O)[O-])cc1', 'COc1cc2nc3c(c(N)c2cc1OC)C1CC(C)=CC(C3)C1', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(Cl)c3C(C1)C2', 'CCC1=CC2Cc3nc4cc(Cl)ccc4c(Cl)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCO)c3C(C1)C2', 'CCC1=CC2Cc3nc4cc(OC)c(OC)cc4c(N)c3C(C1)C2', 'CCC1=CC2Cc3nc4cc(Cl)ccc4c(NCCO)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCl)c3C(C1)C2', 'CC1=CC2Cc3nc4ncccc4c(N)c3C(C1)C2', 'CCC1=CC2Cc3nc4ncccc4c(N)c3C(C1)C2', 'CCC1=CC2Cc3nc4cnccc4c(N)c3C(C1)C2', 'C[C@@H]1CCC[C@H](C)N1S(C)(=O)=O', 'C[C@@H]1CCC[C@H](C)N1S(=O)(=O)c1ccc(Cl)cc1', 'C[C@@H]1CCC[C@H](C)N1S(=O)(=O)c1cc(Cl)ccc1Cl', 'C[C@@H]1CCC[C@H](C)N1S(=O)(=O)c1ccccc1', 'Cc1ccc(S(=O)(=O)N2[C@H](C)CCC[C@@H]2C)cc1', 'C[C@@H]1CCC[C@H](C)N1S(=O)(=O)c1ccc(C(C)(C)C)cc1', 'C[C@@H]1CCC[C@H](C)N1S(=O)(=O)c1ccccc1[N+](=O)[O-]', 'C[C@@H]1CCC[C@H](C)N1S(=O)(=O)c1cccc([N+](=O)[O-])c1', 'C[C@@H]1CCC[C@H](C)N1S(=O)(=O)c1ccc([N+](=O)[O-])cc1', 'CCN(CCCCCCNC1=CC(=O)C(NCCCCCCNC(=O)CCCCC2CCSS2)=CC1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(NCCCNC(=O)CCCCC2CCSS2)=CC1=O)Cc1ccccc1OC', 'CCN(CCCNC1=CC(=O)C(NCCCCCCNC(=O)CCCCC2CCSS2)=CC1=O)Cc1ccccc1OC', 'CCN(CCCNC1=CC(=O)C(NCCCNC(=O)CCCCC2CCSS2)=CC1=O)Cc1ccccc1OC', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(CCCN1CCCCC1)C2', 'O=S(=O)(/N=C(\\NCCCCNc1c2c(nc3ccccc13)CCCC2)N1CC(c2ccccc2)C(c2ccc(Cl)cc2)=N1)c1ccc(Cl)cc1', 'O=S(=O)(/N=C(\\NCCCCCCCNc1c2c(nc3ccccc13)CCCC2)N1CC(c2ccccc2)C(c2ccc(Cl)cc2)=N1)c1ccc(Cl)cc1', 'CCc1c(C(=O)NCCCCNc2c3c(nc4ccccc24)CCCC3)nc(-c2ccccc2Cl)n1-c1ccc(Cl)cc1', 'CCc1c(C(=O)NCCCCCCCNc2c3c(nc4ccccc24)CCCC3)nc(-c2ccccc2Cl)n1-c1ccc(Cl)cc1', 'C[n+]1c(/C=C/c2ccc(F)cc2)ccc2ccccc21.[I-]', 'CCSc1nc2sc3c(c2c(=O)n1-c1ccccc1)CCN(Cc1ccccc1)C3', 'Cc1ccc2c(c1)c1c(n2CCc2ccc(C)nc2)CCN(C)C1', 'COc1cc2occ(-c3ccc(N4CCCC4)cc3)c(=O)c2cc1OC', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(/N=C/c1ccc(OC)cc1)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(/N=C/c1ccc(O)c(OCC)c1)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(/N=C/c1cc(OC)cc(OC)c1)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(/N=C/c1ccc(OC)cc1OC)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(/N=C/c1ccc(C#N)cc1)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(/N=C/C=C/c1ccccc1)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(/N=C/C(=C/c1ccccc1)CCCCC)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(/N=C/C(C)=C/c1ccccc1)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(/N=C/C=C/c1ccc(N(C)C)cc1)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(/N=C/c1ccccc1OS(=O)(=O)[O-])c1ccc(=O)[nH]c1C2.[Na+]', 'C=CCOc1ccc(/C=N/[C@@]23CC(C)=C[C@@H](Cc4[nH]c(=O)ccc42)/C3=C\\C)cc1', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(/N=C/c1ccc(O)c(OC)c1)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(/N=C/c1ccc(N(C)C)cc1)c1ccc(=O)[nH]c1C2', 'COc1ccc(/C=C2\\Oc3ccccc3C2=O)cc1', 'CC(C)(C)c1cc[n+](CCCCCCCCCCCC[n+]2ccc(C(C)(C)C)cc2)cc1.[Br-].[Br-]', 'NC(=O)c1cc[n+](C/C=C/C[n+]2ccc(/C=N/O)c(F)c2)cc1.[Br-].[Br-]', 'Cn1ccc2nc3ccccc3c-2c1.I', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(CN(C)Cc3ccccc3)cc1)C2', 'COc1ccc2c(c1)C/C(=C\\c1ccc(CN(C)Cc3ccccc3)cc1)C2=O', 'Clc1cccc(Cl)c1CO/N=C/c1cc[n+](CCC[n+]2ccc(/C=N/OCc3c(Cl)cccc3Cl)cc2)cc1.[Br-].[Br-]', 'COc1ccc2c(c1)C(=O)/C(=C/c1ccc(CN(C)Cc3ccccc3)cc1)C2', 'CCN(Cc1ccccc1)Cc1ccc(/C=C2\\Cc3ccc(OC)cc3C2=O)cc1', 'CN(Cc1ccccc1)Cc1ccc(/C=C2\\Cc3ccccc3C2=O)cc1', 'CCN(CC)CCOc1ccc2c(c1)C/C(=C\\c1ccc(CN(C)Cc3ccccc3)cc1)C2=O', 'CCN(CC)CCCOc1ccc2c(c1)C/C(=C\\c1ccc(CN(C)Cc3ccccc3)cc1)C2=O', 'CCN(CC)CCCCOc1ccc2c(c1)C/C(=C\\c1ccc(CN(C)Cc3ccccc3)cc1)C2=O', 'CCN(CC)CCOc1ccc2c(c1)C(=O)/C(=C/c1ccc(CN(C)Cc3ccccc3)cc1)C2', 'CCN(CC)CCCOc1ccc2c(c1)C(=O)/C(=C/c1ccc(CN(C)Cc3ccccc3)cc1)C2', 'CCN(CC)CCCCOc1ccc2c(c1)C(=O)/C(=C/c1ccc(CN(C)Cc3ccccc3)cc1)C2', 'CN(Cc1ccccc1)Cc1ccc(/C=C2\\Cc3ccc(OCCCCN4CCOCC4)cc3C2=O)cc1', 'CN(Cc1ccccc1)Cc1ccc(/C=C2\\Cc3ccc(OCCCCN4CCCCC4)cc3C2=O)cc1', 'CCN(CC)CCCCCOc1ccc2c(c1)C(=O)/C(=C/c1ccc(CN(C)Cc3ccccc3)cc1)C2', 'CN(Cc1ccccc1)Cc1ccc(/C=C2\\Cc3ccc(OCCCCCN4CCCCC4)cc3C2=O)cc1', 'CN(Cc1ccccc1)Cc1ccc(/C=C2\\Cc3ccc(OCCCCCN4CCCCC4)cc3C2=O)cc1', 'CCN(CC)CCCCCCOc1ccc2c(c1)C(=O)/C(=C/c1ccc(CN(C)Cc3ccccc3)cc1)C2', 'CCN(CC)CCCCCCCOc1ccc2c(c1)C(=O)/C(=C/c1ccc(CN(C)Cc3ccccc3)cc1)C2', 'COc1ccc2c(c1)CC/C(=C\\c1ccc(CN(C)Cc3ccccc3)cc1)C2=O', 'COc1ccc2c(c1)C(=O)/C(=C/c1ccc(CN(C)Cc3ccccc3)cc1)CC2', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(CN(C)Cc3ccccc3)cc1)CC2', 'CCN(CC)CCCCCOc1ccc2c(c1)CC/C(=C\\c1ccc(CN(C)Cc3ccccc3)cc1)C2=O', 'CN(Cc1ccccc1)Cc1ccc(/C=C2\\CCc3cc(OCCCCCN4CCCCC4)ccc3C2=O)cc1', 'CN(Cc1ccccc1)Cc1ccc(/C=C2\\CCc3cc(OCCCCCN4CCCCC4)ccc3C2=O)cc1', 'CN(Cc1ccccc1)Cc1ccc(/C=C2\\CCc3cc(OCCCCCN4CCCCC4)ccc3C2=O)cc1', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@]1(C)N2C', 'NC(=O)N1c2ccccc2Sc2ccccc21', 'CNC(=O)N1c2ccccc2Sc2ccccc21', 'CO[P@]1(=O)OC[C@H]2COC(=O)C2=C(C)O1', 'CO[P@]1(=O)OC[C@H]2COC(=O)C2=C(C)O1', 'CO[P@]1(=O)OC[C@H]2COC(=O)C2=C(C)O1', 'CO[P@]1(=O)OC[C@H]2COC(=O)C2=C(C)O1', 'CO[P@@]1(=O)CC[C@H]2COC(=O)C2=C(C)O1', 'CO[P@@]1(=O)CC[C@H]2COC(=O)C2=C(C)O1', 'CO[P@]1(=O)CC[C@H]2COC(=O)C2=C(C)O1', 'CO[P@]1(=O)CC[C@H]2COC(=O)C2=C(C)O1', 'CO[P@]1(=O)CC[C@H]2COC(=O)C2=C(C)O1', 'COC(=O)C1=C(C)O[P@@](=O)(OC)CC[C@H]1COCc1ccccc1', 'COC(=O)C1=C(C)O[P@](=O)(OC)CC[C@H]1COCc1ccccc1', 'COC(=O)C1=C(C)OP(=O)(OC)CCC1', 'COC(=O)C1=C(C)OP(=O)(OC)CCC1', 'COC(=O)C1=C(C)OP(=O)(OC)CCC1', 'COC(=O)C1=C(C)OP(=O)(OC)CCC1', 'COC(=O)C1=C(C)OP(=O)(OC)CC1', 'COP(=O)(OC)O/C(C)=C1\\CCOC1=O', 'CCN(CC)CCNC(=O)N1c2ccccc2Sc2ccccc21.Cl', 'CC(C)N(CCNC(=O)N1c2ccccc2Sc2ccccc21)C(C)C', 'O=C(NCCN1CCCC1)N1c2ccccc2Sc2ccccc21', 'O=C(NCCN1CCCCC1)N1c2ccccc2Sc2ccccc21', 'CS(=O)(=O)[O-].CS(=O)(=O)[O-].O/N=C/c1cc[n+](COC[n+]2ccc(/C=N/O)cc2)cc1', 'CS(=O)(=O)[O-].CS(=O)(=O)[O-].O/N=C/c1cc[n+](COC[n+]2ccc(/C=N/O)cc2)cc1', 'CS(=O)(=O)[O-].CS(=O)(=O)[O-].O/N=C/c1cc[n+](COC[n+]2ccc(/C=N/O)cc2)cc1', 'CS(=O)(=O)[O-].CS(=O)(=O)[O-].O/N=C/c1cc[n+](COC[n+]2ccc(/C=N/O)cc2)cc1', 'CS(=O)(=O)[O-].CS(=O)(=O)[O-].O/N=C/c1cc[n+](COC[n+]2ccc(/C=N/O)cc2)cc1', 'CS(=O)(=O)[O-].CS(=O)(=O)[O-].O/N=C/c1cc[n+](COC[n+]2ccc(/C=N/O)cc2)cc1', '[Br-].[Br-].c1ccc2c(c1)ccc[n+]2CCCCCC[n+]1cccc2ccccc21', 'COc1cc2cc(C(=O)CCc3cc[n+](Cc4ccc([N+](=O)[O-])o4)cc3)sc2cc1OC.[Br-]', 'COc1cc2cc(C(=O)CCc3cc[n+](Cc4ccc([N+](=O)[O-])o4)cc3)sc2cc1OC.[Br-]', 'CC[N+](C)(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCC[N+](C)(CC)Cc1ccccc1[N+](=O)[O-])Cc1ccccc1[N+](=O)[O-].[I-].[I-]', 'CC[N+](C)(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCC[N+](C)(CC)Cc1ccccc1[N+](=O)[O-])Cc1ccccc1[N+](=O)[O-].[I-].[I-]', 'CC[N+](C)(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCC[N+](C)(CC)Cc1ccccc1[N+](=O)[O-])Cc1ccccc1[N+](=O)[O-].[I-].[I-]', 'CC[N+](C)(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCC[N+](C)(CC)Cc1ccccc1[N+](=O)[O-])Cc1ccccc1[N+](=O)[O-].[I-].[I-]', 'COc1cc2c(cc1OC)SC(C(=O)CCc1cc[n+](Cc3coc([N+](=O)[O-])c3)cc1)C2.[I-]', 'COc1cc2c(cc1OC)SC(C(=O)CCc1cc[n+](Cc3coc([N+](=O)[O-])c3)cc1)C2.[I-]', 'COc1cc2cc(C(=O)C3CCC[N+](C)(Cc4ccc([N+](=O)[O-])o4)CC3)sc2cc1OC.[I-]', 'COc1cc2cc(C(=O)C3CCC[N+](C)(Cc4ccc([N+](=O)[O-])o4)CC3)sc2cc1OC.[I-]', 'CCN(CCCCCCNc1ccc(CCc2ccc(NCCCCCCN(CC)Cc3ccccc3OC)cc2)cc1)Cc1ccccc1OC', 'COc1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)C2N3C)cc1', 'COc1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(C)C2N3C)cc1', 'CCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2C', 'CCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2C', 'CC(C)NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2C', 'CC(C)NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2C', 'CCN(CC)C(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2C', 'CCN(CC)C(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2C', 'CN(C(=O)NCc1ccccc1)C(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2C', 'CN(C(=O)NCc1ccccc1)C(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2C', 'CN1CC[C@@]2(C)c3cc(OC(=O)NCc4ccccc4)ccc3N(C)C12', 'CN1CC[C@@]2(C)c3cc(OC(=O)NCc4ccccc4)ccc3N(C)C12', 'CN(C)C(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2C', 'CN(C)C(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2C', 'CCCCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2C', 'CCCCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)C1N2C', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccc(Cl)cc4)ccc3N(C)C12', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccc(Cl)cc4)ccc3N(C)C12', 'CCCCCCOC(=O)[C@H](CCC(=O)NCCC1CCN(Cc2ccccc2)CC1)NC(=O)OC(C)(C)C', 'CCCCCCOC(=O)[C@H](CCC(=O)NCCC1CCN(Cc2ccccc2)CC1)NC(=O)OCc1ccccc1', 'CCCCCCOC(=O)[C@H](CCC(=O)NCCC1CCN(Cc2ccccc2)CC1)NC(=O)Cc1cc2ccc(Cl)cc2s1', 'CCCCCCOC(=O)[C@H](CCC(=O)NCCC1CCN(Cc2ccccc2)CC1)NC(=O)c1ccccc1', 'CCCCCCOC(=O)[C@H](CCC(=O)NCCC1CCN(Cc2ccccc2)CC1)NC(=O)c1cccs1', 'CCCCCCOC(=O)[C@H](CCC(=O)NCCC1CCN(Cc2ccccc2)CC1)NC(=O)c1cc(Br)cs1', 'CCCCCCOC(=O)[C@H](CCC(=O)NCCC1CCN(Cc2ccccc2)CC1)NC(=O)c1cc2ccccc2s1', 'CCCCCCOC(=O)[C@H](CCC(=O)NCCC1CCN(Cc2ccccc2)CC1)NC(=O)c1cc2cccnc2s1', 'CCCCCCOC(=O)[C@H](CCC(=O)NCCC1CCN(Cc2ccccc2)CC1)NC(=O)c1cc2ccsc2s1', '[Br-].[Br-].c1cc[n+](Cc2ccc3ccc(C[n+]4ccccc4)cc3c2)cc1', '[Br-].[Br-].c1cc[n+](Cc2ccc3ccc(C[n+]4ccccc4)cc3c2)cc1', '[Br-].[Br-].c1cc[n+](Cc2ccc3ccc(C[n+]4ccccc4)cc3c2)cc1', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(CCc4ccccc4)[C@@H]2N3C)cc1', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN(CCc4ccccc4)[C@@H]2N3C)cc1', 'CN1c2ccc(OC(=O)NC3CCCCC3)cc2[C@]2(C)CCN(CCc3ccccc3)[C@H]12', 'CN1c2ccc(OC(=O)NCC3CCCCC3)cc2[C@]2(C)CCN(CCc3ccccc3)[C@H]12', 'CN1c2ccc(OC(=O)NCC3CCCCC3)cc2[C@]2(C)CCN(CCc3ccccc3)[C@H]12', 'CC(CC(=O)C[C@@H](C)[C@H]1C[C@H](O)[C@@]2(C)C3=C(C(=O)C[C@]12C)[C@@]1(C)CCC(=O)C(C)(C)[C@@H]1CC3)C(=O)O', 'COC(=O)C(C)CC(=O)C[C@@H](C)[C@H]1C[C@H](O)[C@@]2(C)C3=C(C(=O)C[C@]12C)[C@@]1(C)CCC(=O)C(C)(C)[C@@H]1C[C@@H]3O', 'CC(CC(=O)C[C@@H](C)[C@H]1CC(=O)[C@@]2(C)C3=C(C(=O)C[C@]12C)[C@@]1(C)CC[C@H](O)C(C)(C)[C@@H]1C[C@@H]3O)C(=O)O', 'CC(C)=CCC/C(C)=C/Cc1cc(C2CC(=O)c3c(O)cc(O)cc3O2)c(O)cc1O', 'C[C@H](CC[C@@H](O)C(C)(C)O)[C@H]1CC[C@@]2(C)C3=CC[C@H]4C(C)(C)C(=O)CC[C@]4(C)C3=CC[C@]12C', 'CCOC(=O)c1cc2c(N)c3c(nc2nc1C)CCCC3', 'CCOC(=O)c1cc2c(N)c3c(nc2nc1C)CCCC3', 'CCOC(=O)c1cc2c(N)c3c(nc2nc1C)CCCC3', 'CCCCCCNC(=O)Oc1ccc2c(c1)C1CCN(CCC)C1C2', 'CCCCCCNC(=O)Oc1cccc(CN(CC)CCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'c1ccc2c(NCCCCCCCCCNc3ccncc3)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCCCCCCCCCNc3ccncc3)c3c(nc2c1)CCCC3', 'c1ccc(NCCCCCCCCCCCCNc2c3c(nc4ccccc24)CCCC3)cc1', 'c1ccc2c(NCCCCCCCNc3c4c(nc5ccccc35)CCCC4)ccnc2c1', 'Cc1ccc2nc3c(c(Nc4ccccc4)c2c1)CCCC3', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)CCCN2', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)CCCN2', 'CCCCCCNC(=O)Oc1cccc(CN(CCC)CCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'c1ccc2c(NCCCCCCCCNc3c4c(nc5ccccc35)CCCC4)ccnc2c1', 'c1ccc2c(NCCCCCCCCCNc3c4c(nc5ccccc35)CCCC4)ccnc2c1', 'Cc1ccc2c(Nc3ccccc3)c3c(nc2c1)CCCC3', 'Clc1ccc2c(Nc3ccccc3)c3c(nc2c1)CCCC3', 'O=[N+]([O-])c1ccc2nc3c(c(Nc4ccccc4)c2c1)CCCC3', 'O=[N+]([O-])c1ccc2c(Nc3ccccc3)c3c(nc2c1)CCCC3', 'CN1CCCc2cc(C(=O)CCC3CCN(Cc4ccccc4)CC3)ccc21', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)CCCN2c1ccccc1', 'CCCCCCNC(=O)Oc1cccc(CN(CCCOc2ccc3c(=O)c4ccccc4oc3c2)C(C)C)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(c2)O/C(=C\\c2cc(O)c(O)c(O)c2)C3=O)c1', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)N(c1ccccc1)CCC2', 'Cc1cc(-c2ccccc2)nnc1NCCN1CCCCC1', 'Cc1cc(-c2ccc(Cl)cc2)nnc1NCCN1CCCCC1', 'Cc1cc(-c2ccc3c(c2)OCO3)nnc1NCCN1CCCCC1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)O/C(=C\\c2cc(O)c(O)c(O)c2)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)O/C(=C\\c2ccc4ccccc4c2)C3=O)c1', 'COC(=O)c1cc2c(N)c3c(nc2nc1C)CCC3', 'CCN1CCCC1CNc1ccc(-c2ccccc2)nn1', 'c1ccc(-c2ccc(NCCN3CCc4ccccc4C3)nn2)cc1', 'O=C1c2ccccc2CC1CC1CCN(c2ccccc2)CC1', 'O=C1c2ccccc2CCC1CC1CCN(c2ccccc2)CC1', 'O=C1c2ccccc2CCCC1CC1CCN(c2ccccc2)CC1', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(c3ccccc3)CC1)C2', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)C/C(=C\\c2cc(O)c(O)c(O)c2)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)OC/C(=C/c2cc(O)c(O)c(O)c2)C3=O)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc3c(c2)CC/C(=C/c2cc(O)c(O)c(O)c2)C3=O)c1', 'COC(=O)c1cc2c(N)c3c(nc2nc1C)CCCC3', 'COC(=O)c1cc2c(N)c3c(nc2nc1C)CCCCC3', 'COC(=O)c1cc2c(N)c3c(nc2nc1C)CCCCCC3', 'CCOC(=O)c1cc2c(N)c3c(nc2nc1C)CCC3', 'CCOC(=O)c1cc2c(N)c3c(nc2nc1C)CCCCC3', 'c1ccc(-c2ccc(NCCCN3CCc4ccccc4C3)nn2)cc1', 'c1ccc(-c2ccc(NCCCCCN3Cc4ccccc4CN3)nn2)cc1', 'CN(CCCNc1ccc(-c2ccccc2)nn1)Cc1ccccc1', 'COc1cc2c(cc1OC)C(O)C(CC1CCN(c3ccccc3)CC1)C2', 'COc1cc2c(cc1OC)CC(CC1CCN(c3ccccc3)CC1)=C2', 'O=C(CCc1c[nH]c2ccccc12)NCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(CCc1c[nH]c2ccccc12)NCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'CCOC(=O)c1cc2c(N)c3c(nc2nc1C)CCCCCC3', 'O=C(Nc1ccccc1)c1ccc2c(c1)C(=O)N(CCC1CCN(Cc3ccccc3)CC1)C2=O', 'CN(CCCCNc1ccc(-c2ccccc2)nn1)Cc1ccccc1', 'CN(CCCCCNc1ccc(-c2ccccc2)nn1)Cc1ccccc1', 'c1ccc(CN2CCC(Nc3ccc(-c4ccccc4)nn3)CC2)cc1', 'c1ccc(CN2CCC(CNc3ccc(-c4ccccc4)nn3)CC2)cc1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'O=C(CCc1c[nH]c2ccccc12)NCCCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCc1c[nH]c2ccccc12)NCCCCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'c1ccc(-c2ccc(NCN3CCCCC3)nc2)cc1', 'c1ccc(-c2cnc(NCN3CCCCC3)nc2)cc1', 'c1ccc(-c2ncc(NCN3CCCCC3)nn2)cc1', 'c1ccc(-c2nnc(NCN3CCCCC3)s2)cc1', 'c1ccc(-c2cnc(NCN3CCCCC3)nn2)cc1', 'c1ccc(C2=NNC(NCN3CCCCC3)=NC2)cc1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc(C(=O)c3ccccc3)cc2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc(Cc3ccccc3)cc2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccc3ccccc3c2)c1', 'CNC(=O)Oc1cccc(CN(C)CCCOc2ccccc2)c1', 'CN(C)C(=O)Oc1ccc(-c2cn3c(Cl)cccc3[n+]2C)cc1.[Cl-]', 'Cc1cccc2n1cc(-c1ccc(OC(=O)N(C)C)cc1)[n+]2C.[Cl-]', 'CN(C)C(=O)Oc1ccc(-c2cn3ccccc3[n+]2C)cc1.Cc1ccc(S(=O)(=O)[O-])cc1', 'CN(C)C(=O)Oc1ccc(-c2cn3c(C(F)(F)F)cccc3[n+]2C)cc1.[Cl-]', 'COc1cccc2n1cc(-c1ccc(OC(=O)N(C)C)cc1)[n+]2C.[Cl-]', 'CN(C)C(=O)Oc1ccc(-c2cn3c(NC=O)cccc3[n+]2C)cc1.[Cl-]', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccccc1', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccccc1', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)OCO2', 'CCn1c(C)nc2cc(C(=O)CCC3CCN(Cc4ccccc4)CC3)ccc21', 'CCn1c(C)nc2cc(C(=O)CCC3CCN(Cc4ccccc4)CC3)ccc21', 'Cc1nc2ccc(C(=O)CCC3CCN(Cc4ccccc4)CC3)cc2s1', 'Cc1nc2ccc(C(=O)CCC3CCN(Cc4ccccc4)CC3)cc2s1', 'CNC(=O)Oc1cccc(CN(C)CC#CCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CC(=O)Nc1cccc2n1cc(-c1ccc(OC(=O)N(C)C)cc1)[n+]2C.[Cl-]', 'CC(=O)N(C)c1cccc2n1cc(-c1ccc(OC(=O)N(C)C)cc1)[n+]2C.[Cl-]', 'CN(C)C(=O)Oc1ccc(-c2cn3c(NC(=O)c4ccccc4)cccc3[n+]2C)cc1.[Cl-]', 'CNC(=O)Nc1cccc2n1cc(-c1ccc(OC(=O)N(C)C)cc1)[n+]2C.[Cl-]', 'COC(=O)Nc1cccc2n1cc(-c1ccc(OC(=O)N(C)C)cc1)[n+]2C.[Cl-]', 'Cc1nc2cc(C(=O)CCC3CCN(Cc4ccccc4)CC3)ccc2[nH]1', 'Cc1nc2cc(C(=O)CCC3CCN(Cc4ccccc4)CC3)ccc2[nH]1', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2[nH]c(-c3ccccc3)nc2c1', 'CN(C)C(=O)Nc1cccc2n1cc(-c1ccc(OC(=O)N(C)C)cc1)[n+]2C.[Cl-]', 'CNC(=O)Oc1ccc(-c2cn3ccccc3[n+]2C)cc1.[I-]', 'CNC(=O)Oc1ccc(-c2cn3c(C)cccc3[n+]2C)cc1.[I-]', 'CNC(=O)Oc1ccc(-c2cn3c(NC(C)=O)cccc3[n+]2C)cc1.[I-]', 'CN(C)C(=O)Oc1cccc(-c2cn3ccccc3[n+]2C)c1.[Cl-]', 'Cc1cccc2n1cc(-c1cccc(OC(=O)N(C)C)c1)[n+]2C.[Cl-]', 'O=C1CNc2cc3onc(CCC4CCN(Cc5ccccc5)CC4)c3cc2C1', 'O=C(CCOCCc1c[nH]c2ccccc12)NCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'CN(C)C(=O)Oc1cccc(-c2cn3c(Cl)cccc3[n+]2C)c1.[Cl-]', 'CN(C)C(=O)Oc1cccc(-c2cn3c(C(F)(F)F)cccc3[n+]2C)c1.[Cl-]', 'COc1cccc2n1cc(-c1cccc(OC(=O)N(C)C)c1)[n+]2C.[Cl-]', 'CC(=O)Nc1cccc2n1cc(-c1cccc(OC(=O)N(C)C)c1)[n+]2C.[Cl-]', 'CNC(=O)Oc1cccc(-c2cn3ccccc3[n+]2C)c1.[I-]', 'c1ccc(CN2CCC(CNc3nsc4ccccc34)CC2)cc1', 'c1ccc(CN2CCC(CNc3n[nH]c4ccccc34)CC2)cc1', 'CCCCCCNC(=O)Oc1cccc(CN(C)CCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CCCCCCNC(=O)Oc1cccc(CN(C)CCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CN(CCOCCNC(=S)NC(=O)c1ccc2c(c1)C(=O)c1ccccc1-2)Cc1ccccc1', 'CN(CCOCCNC(=S)NC(=O)c1ccc2c(c1)C(=O)c1ccccc1C2=O)Cc1ccccc1', 'O=C(CCOCCc1c[nH]c2ccccc12)NCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCOCCc1c[nH]c2ccccc12)NCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'CNC(=O)Oc1cccc(-c2cn3c(NC(C)=O)cccc3[n+]2C)c1.[I-]', 'Cc1cn2cc(C(N)=O)ccc2[n+]1C.[I-]', 'C[n+]1c(-c2ccccc2)cn2cc(C(N)=O)ccc21.[I-]', 'COC(=O)Nc1ccc2n(c1)cc(-c1ccccc1)[n+]2C.[Cl-]', 'COc1ccc2c(c1OC)C(=O)CC2CCC1CCN(Cc2ccccc2)CC1', 'CCCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1C2', 'CCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1C2', 'CCCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CC)[C@@H]1C2', 'CN(CCOCCNC(=S)Nc1ccccn1)Cc1ccccc1', 'CCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CC)[C@@H]1C2', 'CCN1CC[C@@]2(C)c3cc(OC(=O)NC(C)(C)C)ccc3C[C@@H]12', 'CCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CC)[C@@H]1C2', 'CCN1CC[C@@]2(C)c3cc(OC(=O)NC)ccc3C[C@@H]12', 'CCN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3C[C@@H]12', 'CCCCCCNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)c4cccnc4oc3c2)c1', 'NCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'NCCCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'C/C(=N\\O)c1ccc(Cl)c(NC(=O)c2c(-c3ccccc3Cl)noc2C)c1', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)CCN2c1ccccc1', 'CCN1CC[C@@]2(C)c3cc(OC(=O)NC(C)c4ccccc4)ccc3C[C@@H]12', 'CCN1CC[C@@]2(C)c3cc(OC(=O)NCc4ccccc4)ccc3C[C@@H]12', 'CCCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CCC)[C@@H]1C2', 'CCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CCC)[C@@H]1C2', 'CCCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(Cc3ccccc3)[C@@H]1C2', 'CCCCCCNC(=O)Oc1cccc(CN(C)CCCOc2ccc3ccc(=O)oc3c2)c1', 'CCCCNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)ccoc3c2)c1', 'NCCCCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'NCCCCCCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'CN(C)CCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'CN(C)CCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'CN(C)CCCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'c1ccc2c(NCCCCCCNc3c4c(nc5ncccc35)CCCCC4)c3c(nc2c1)CCCCC3', 'CCCc1cc(N2CCC=N2)ccc1C(=O)O/N=C(\\N)c1ccc(C(F)(F)F)cc1', 'CCCCCCCNC(=O)Oc1ccc2c(c1)[C@@H]1CCN(CC)[C@@H]1C2', 'CCCCCCNC(=O)Oc1ccc2c(c1)[C@@H]1CCN(CC)[C@@H]1C2', 'CCCCNC(=O)Oc1ccc2c(c1)C1CCN(CC)C1C2', 'CCCNC(=O)Oc1ccc2c(c1)C1CCN(CC)C1C2', 'CCN1CCC2c3cc(OC(=O)NC)ccc3CC21', 'CCCCCCCNC(=O)Oc1ccc2c(c1)C1CCN(CCC)C1C2', 'CCCCCCNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)ccoc3c2)c1', 'CCCCCCNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c(=O)cc(-c4ccccc4)oc3c2)c1', 'CN(C)CCCCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'CN(C)CCCCCCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'c1ccc2c(NCCCCCCCNc3ccncc3)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCCCCCCCNc3ccncc3)c3c(nc2c1)CCCC3', 'Nc1ccc2c(c1)c(-c1ccccc1)[n+](CCCNCCCNCCCNc1c3c(nc4ccccc14)CCCC3)c1cc(N)ccc21.[Cl-]', 'COc1ccccc1CN(C)CCCN(C)CCCNCCC[n+]1c(-c2ccccc2)c2cc(N)ccc2c2ccc(N)cc21.[Cl-]', 'c1ccc(CNc2ccnc(N3CCCC3)n2)cc1', 'c1ccc(CNc2ccnc(N3CCOCC3)n2)cc1', 'c1ccc(CNc2ccnc(N3CCSCC3)n2)cc1', 'CN1CCN(c2nccc(NCc3ccccc3)n2)CC1', 'CC1CCN(c2nccc(NCc3ccccc3)n2)CC1', 'c1ccc(CCNc2ccnc(N3CCCC3)n2)cc1', 'c1ccc(CCNc2ccnc(N3CCOCC3)n2)cc1', 'c1ccc(CCNc2ccnc(N3CCSCC3)n2)cc1', 'CN1CCN(c2nccc(NCCc3ccccc3)n2)CC1', 'CC1CCN(c2nccc(NCCc3ccccc3)n2)CC1', 'c1ccc2c(CNc3ccnc(N4CCCC4)n3)cccc2c1', 'c1ccc2c(CNc3ccnc(N4CCOCC4)n3)cccc2c1', 'c1ccc2c(CNc3ccnc(N4CCSCC4)n3)cccc2c1', 'CN1CCN(c2nccc(NCc3cccc4ccccc34)n2)CC1', 'CN1CCN(c2nccc(NCc3cccc4ccccc34)n2)CC1', 'NC1=C2CCCCC2=NC2CCCCC12', 'Nc1c2c(nc3cccc(F)c13)C1CC(C2)C1', 'Nc1c2c(nc3ccc(O)cc13)CCCC2', 'C=C1c2c(Cl)cccc2N=C2CCCCCN12', 'C=C1c2ccc(Cl)cc2N=C2CCCCCN12', 'C=C1c2cc(Cl)ccc2N=C2CCCCCN12', 'C=C1c2ccccc2N=C2CCCCCN12', 'C=C1c2cccc(Cl)c2N=C2CCCCCN12', 'CC1=C[C@@H]2Cc3nc4cc(Cl)ccc4c(N)c3[C@H](C1)C2', 'CN(C)c1cccc(OC(=O)Nc2ccccc2)c1', 'COc1cc2c(cc1OC)[C@@H]1[C@H](O)[C@@H](OC)C=C3CCN(C2)[C@H]31', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CC[N+](C)(CCc4ccccc4)C2N3C)cc1.[Br-]', 'COS(=O)(=O)[O-].C[N+](C)(C)c1cccc(OC(=O)Nc2ccccc2)c1', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CC[N+](C)(C)C1N2C.[Br-]', 'CN1c2ccc(OC(=O)Nc3ccccc3)cc2[C@]2(C)CC[N+](C)(C)C12.[Br-]', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CC[N+](C)(C)C1N2C.[Br-]', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CC[N+](C)(C)C2N3C)cc1.[Br-]', 'CN1C(=O)Cc2cc3c(CCC4CCN(Cc5ccccc5)CC4)noc3cc21', 'OC1CCN2Cc3ccccc3N=C12', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(CCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)CC1)C2', 'C#CCN(C)C1CCc2c(OC(=O)N(C)CC)cccc21', 'N#CC(c1ccc2c(c1)OCO2)N1CCOCC1', 'C#CCNC1CCc2c(OC(=O)N(C)C)cccc21', 'O=C(NCCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)c1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)OCCC3', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)OCCO2', 'CC[C@]1(O)CC[C@H]2[C@@H]3CCC4=CCCC[C@@H]4[C@H]3CC[C@@]21C', 'CCN(CC)C(C)CN1c2ccccc2Sc2ccccc21.Cl', 'CCN(CC)C(C)CN1c2ccccc2Sc2ccccc21.Cl', 'CCN(CC)C(C)CN1c2ccccc2Sc2ccccc21.Cl', 'CCN(CC)C(C)CN1c2ccccc2Sc2ccccc21.Cl', 'CCN(CC)C(C)CN1c2ccccc2Sc2ccccc21.Cl', 'CCN(CC)C(C)CN1c2ccccc2Sc2ccccc21.Cl', 'CCN(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1.O=C(O)[C@H](O)[C@@H](O)C(=O)O', 'CCCCOc1ccc(C(=O)CCN2CCCCC2)cc1', 'CN(C)Cc1ccc(CSCC/N=C/NCCSCc2ccc(CN(C)C)o2)o1.Cl', 'Cl.Cl.O.c1ccc(CN2CCC(CCNc3cc4c(nn3)-c3ccccc3CCC4)CC2)cc1', 'Cl.Cl.c1ccc(CN2CCC(CCSc3ccc(-c4ccccc4)nn3)CC2)cc1', 'Cl.Cl.O.c1ccc(CN2CCC(CCOc3ccc(-c4ccccc4)nn3)CC2)cc1', 'Cl.Cl.O.c1ccc(CN2CCC(CCNc3cc4ccc5ccccc5c4nn3)CC2)cc1', 'Cl.Cl.O.c1ccc(CN2CCC(CCNc3cc4c(nn3)-c3ccccc3CC4)CC2)cc1', 'Cl.N#CC(C#N)=C1NCCN1CCNCc1ccc(CN2CC=CCC2)o1', 'COC1CCCN(Cc2ccc(CNCCN3CCNC3=C(C#N)C#N)o2)C1.Cl', 'CNC(NCCNCc1ccc(CN2CCCCC2)o1)=C(C#N)C#N.Cl', 'Cl.N#CC(C#N)=C1NCCN1CCNCc1ccc(CN2CCCCCC2)o1', 'CCN(CC)Cc1ccc(CNCCN2CCNC2=C(C#N)C#N)o1.Cl', 'CC[N+](CC)(CC)CCOc1cc(OCC[N+](CC)(CC)CC)c(C(=O)NCCCCCC(=O)NNc2c3c(nc4ccccc24)CCCC3)cc1OCC[N+](CC)(CC)CC.Cl.[Br-].[Br-].[Br-]', 'CC[N+](CC)(CC)CCOc1ccc(C(=O)NCCCCCC(=O)NNc2c3c(nc4ccccc24)CCCC3)cc1.Cl.[Br-]', 'CC[N+](CC)(CC)CCOc1cc(OCC[N+](CC)(CC)CC)cc(C(=O)NCCCCCC(=O)NNc2c3c(nc4ccccc24)CCCC3)c1.Cl.[Br-].[Br-]', 'Cl.O.c1ccc(CN2CCC(CCN=c3cc4c(n[nH]3)-c3ccccc3C4)CC2)cc1', 'CCc1cc(=NCCC2CCN(Cc3ccccc3)CC2)[nH]nc1-c1ccccc1.Cl.Cl.O', 'Cc1cc(=NCCC2CCN(Cc3ccccc3)CC2)[nH]nc1-c1ccccc1.Cl.Cl.O', 'O=C(CCCN1C(=O)c2ccccc2C1=O)N1CCCCCC1', 'CC(C)(C)[Si](C)(C)O[C@H]1C=C2CCN3Cc4cc5c(cc4[C@H]([C@@H]1OC(=O)c1ccccc1)[C@@H]23)OCO5', 'O=C(O[C@H]1[C@H]2c3cc4c(cc3CN3CCC(=C[C@@H]1O)[C@H]23)OCO4)c1ccccc1', 'CC(=O)O[C@H]1C=C2CCN3Cc4cc5c(cc4[C@H]([C@@H]1OC(=O)c1ccccc1)[C@@H]23)OCO5', 'C[C@H](c1cccc(OC(=O)N(C)C)c1)N1CCCCC1', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2', 'CCNC(=O)Oc1ccc2c(c1)C1(C)COC(O2)O1', 'CN(C)c1ccc(/C=C/c2ccc3ccccc3[n+]2C)cc1.[I-]', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCOC2O3)cc1', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2Cc1ccccc1', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(Cc3ccccc3)[C@@H]12', 'Cc1ccccc1NC(=O)Oc1ccc2c(c1)[C@]1(C)CCN[C@@H]1N2', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(C)[C@@H]1N2C=O', 'CN1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3N(C=O)[C@@H]12', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(Cc3ccccc3)[C@@H]1N2Cc1ccccc1', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)C2(C)COC(O3)O2)cc1', 'CNC(=O)Oc1ccc2c(c1)[C@@]1(C)CCN(C)[C@@H]1N2C', 'C[C@@]12CCN(Cc3ccccc3)[C@@H]1N(Cc1ccccc1)c1ccc(OC(=O)Nc3ccccc3)cc12', 'CNC(=O)Oc1ccc2c(c1)SC[C@@H]1CCN(C)[C@H]21', 'CNC(=O)Oc1ccc2c(c1)CC[C@@H]1CCN(C)[C@H]21', 'CCOC(=O)C1=CN(C)c2cccc(OC(=O)N(C)C)c2C1', 'O=C(Nc1ccccc1Cl)Oc1ccc2c(c1)CCCN2Cc1ccccc1', 'O=C(CCCCCCCCc1ccccc1)c1c(O)cccc1O', 'C/C=C1\\[C@H]2C=C(C)C[C@@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@@]1(N)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@@]1(N)c1ccc(=O)[nH]c1C2', 'COc1ccc2nc3c(c(N)c2c1)CCCC3', 'COc1ccc2nc3c(c(N)c2c1)CCCC3', 'COc1ccc2nc3c(c(N)c2c1)CCCC3', 'COc1ccc2nc3c(c(N)c2c1)CCCC3', 'COc1ccc2nc3c(c(N)c2c1)CCCC3', 'COc1ccc2nc3c(c(N)c2c1)CCCC3', 'COc1ccc2nc3c(c(N)c2c1)CCCC3', 'COc1ccc2nc3c(c(N)c2c1)CCCC3', 'COc1ccc2nc3c(c(N)c2c1)CCCC3', 'COc1ccc2nc3c(c(N)c2c1)CCCC3', 'COc1ccc2nc3c(c(N)c2c1)CCCC3', 'COc1ccc2nc3c(c(N)c2c1)CCCC3', 'COc1ccc2nc3c(c(N)c2c1)CCCC3', 'COc1ccc2nc3c(c(N)c2c1)CCCC3', 'COc1ccc2nc3c(c(N)c2c1)CCCC3', 'CCCCCCCCCCCCCCNc1c2c(nc3ccc(OC)cc13)CCCC2.Cl', 'CCCCCCCCCCCCCNc1c2c(nc3ccc(OC)cc13)CCCC2.Cl', 'CCCCCCCCCCCCNc1c2c(nc3ccc(OC)cc13)CCCC2.Cl', 'CCCCCCCCCNc1c2c(nc3ccc(OC)cc13)CCCC2.Cl', 'CCCCCCCCNc1c2c(nc3ccc(OC)cc13)CCCC2.Cl', 'CCCCCCCNc1c2c(nc3ccc(OC)cc13)CCCC2.Cl', 'CCCCCCNc1c2c(nc3ccc(OC)cc13)CCCC2.Cl', 'CCCCCNc1c2c(nc3ccc(OC)cc13)CCCC2.Cl', 'CCCCNc1c2c(nc3ccc(OC)cc13)CCCC2.Cl', 'CCCNc1c2c(nc3ccc(OC)cc13)CCCC2.Cl', 'CCNc1c2c(nc3ccc(OC)cc13)CCCC2.Cl', 'COc1ccc(C2CN(C)C3(C(=O)Nc4ccc([N+](=O)[O-])cc43)C23Cc2cc(OC)c(OC)cc2C3=O)cc1OC', 'COc1ccc(C2C(c3ccccc3)NC3(C(=O)Nc4ccccc43)C23Cc2cc(OC)c(OC)cc2C3=O)cc1OC', 'COc1cc2c(cc1OC)C(=O)C1(C2)C(c2ccc(F)cc2)C(c2ccccc2)NC12C(=O)Nc1ccc(Cl)cc12', 'COc1ccc(C2C(c3ccccc3)NC3(C(=O)Nc4ccc(Cl)cc43)C23Cc2cc(OC)c(OC)cc2C3=O)cc1OC', 'COc1cc2c(cc1OC)C(=O)C1(C2)C(c2ccccc2Cl)C(c2ccccc2)NC12C(=O)Nc1ccc(Cl)cc12', 'COc1cc2c(cc1OC)C(=O)C1(C2)C(c2ccc(F)cc2)C(c2ccccc2)NC12C(=O)Nc1ccc([N+](=O)[O-])cc12', 'COc1ccc(C2C(c3ccccc3)NC3(C(=O)Nc4ccc([N+](=O)[O-])cc43)C23Cc2cc(OC)c(OC)cc2C3=O)cc1OC', 'COc1cc2c(cc1OC)C(=O)C1(C2)C(c2ccccc2)C(c2ccccc2)NC12C(=O)Nc1ccc([N+](=O)[O-])cc12', 'COc1cc2c(cc1OC)C(=O)C1(C2)C(c2ccccc2Cl)C(c2ccccc2)NC12C(=O)Nc1ccc([N+](=O)[O-])cc12', 'COc1cc2c(cc1OC)C(=O)C1(C2)C(c2ccc(F)cc2)CN(C)C12C(=O)Nc1ccccc12', 'COc1ccc(C2CN(C)C3(C(=O)Nc4ccccc43)C23Cc2cc(OC)c(OC)cc2C3=O)cc1OC', 'COc1cc2c(cc1OC)C(=O)C1(C2)C(c2ccccc2)CN(C)C12C(=O)Nc1ccccc12', 'COc1cc2c(cc1OC)C(=O)C1(C2)C(c2ccccc2Cl)CN(C)C12C(=O)Nc1ccccc12', 'COc1cc2c(cc1OC)C(=O)C1(C2)C(c2ccc(F)cc2)CN(C)C12C(=O)Nc1ccc(Cl)cc12', 'COc1ccc(C2CN(C)C3(C(=O)Nc4ccc(Cl)cc43)C23Cc2cc(OC)c(OC)cc2C3=O)cc1OC', 'COc1cc2c(cc1OC)C(=O)C1(C2)C(c2ccccc2)CN(C)C12C(=O)Nc1ccc(Cl)cc12', 'COc1cc2c(cc1OC)C(=O)C1(C2)C(c2ccccc2Cl)CN(C)C12C(=O)Nc1ccc(Cl)cc12', 'COc1cc2c(cc1OC)C(=O)C1(C2)C(c2ccc(F)cc2)CN(C)C12C(=O)Nc1ccc([N+](=O)[O-])cc12', 'COc1cc2c(cc1OC)C(=O)C1(C2)C(c2ccccc2)CN(C)C12C(=O)Nc1ccc([N+](=O)[O-])cc12', 'COc1cc2c(cc1OC)C(=O)C1(C2)C(c2ccccc2Cl)CN(C)C12C(=O)Nc1ccc([N+](=O)[O-])cc12', 'COc1cc2c(cc1OC)C(=O)C1(C2)C(c2ccc(F)cc2)C(c2ccccc2)NC12C(=O)Nc1ccccc12', 'COc1cc2c(cc1OC)C(=O)C1(C2)C(c2ccccc2)C(c2ccccc2)NC12C(=O)Nc1ccccc12', 'COc1cc2c(cc1OC)C(=O)C1(C2)C(c2ccccc2Cl)C(c2ccccc2)NC12C(=O)Nc1ccccc12', 'COc1cc2c(cc1OC)C(=O)C1(C2)C(c2ccccc2)C(c2ccccc2)NC12C(=O)Nc1ccc(Cl)cc12', 'C[n+]1c(/C=C/c2ccc(O)cc2)ccc2ccccc21.[I-]', 'Cc1ccc(C(=O)Nc2cc3ccccc3oc2=O)cc1', 'COc1cccc(C(=O)N2c3ccccc3Sc3ccccc32)c1', 'COc1ccc(NC(S)=NCCN2CCC(C)CC2)cc1Cl', 'COc1ccc(C(C#N)N2CCCCC2)cc1OC', 'Clc1ccc(CNC2CCN(Cc3ccccc3)CC2)cc1', 'N#CC(c1ccc(O)cc1)N1CCCCC1', 'O=C1/C(=C/c2ccc(O)cc2)COc2ccccc21', 'Clc1ccc2c(c1)Nc1ccccc1S2', 'Fc1ccc(CNC2CCN(Cc3ccccc3)CC2)cc1', 'O=C(COc1ccc(Cl)cc1Cl)Nc1ccc(CN2CCCCC2)cc1', 'Cc1cc(N2CCCCC2)nc2ccccc12', 'CCCCCCCNc1c2c(nc3ccc(OC)cc13)CCCC2', 'CCCCCCNc1c2c(nc3ccc(OC)cc13)CCCC2', 'CNc1nc(NCc2ccccc2)c2ccccc2n1', 'CC(C)(C)c1cc[n+](C[n+]2ccc(C(C)(C)C)cc2)cc1.[Br-].[Br-]', 'CC(C)(C)c1cc[n+](CC[n+]2ccc(C(C)(C)C)cc2)cc1.[Br-].[Br-]', 'CC(C)(C)c1cc[n+](CCC[n+]2ccc(C(C)(C)C)cc2)cc1.[Br-].[Br-]', 'CC(C)(C)c1cc[n+](CCCC[n+]2ccc(C(C)(C)C)cc2)cc1.[Br-].[Br-]', 'CC(C)(C)c1cc[n+](CCCCC[n+]2ccc(C(C)(C)C)cc2)cc1.[Br-].[Br-]', 'CC(C)(C)c1cc[n+](CCCCCC[n+]2ccc(C(C)(C)C)cc2)cc1.[Br-].[Br-]', 'CC(C)(C)c1cc[n+](CCCCCCC[n+]2ccc(C(C)(C)C)cc2)cc1.[Br-].[Br-]', 'CC(C)(C)c1cc[n+](CCCCCCCC[n+]2ccc(C(C)(C)C)cc2)cc1.[Br-].[Br-]', 'CC(C)(C)c1cc[n+](CCCCCCCC[n+]2ccc(C(C)(C)C)cc2)cc1.[Br-].[Br-]', 'CC(C)(C)c1cc[n+](CCCCCCCCC[n+]2ccc(C(C)(C)C)cc2)cc1.[Br-].[Br-]', 'CC(C)(C)c1cc[n+](CCCCCCCCCC[n+]2ccc(C(C)(C)C)cc2)cc1.[Br-].[Br-]', 'CC(C)(C)c1cc[n+](CCCCCCCCCCC[n+]2ccc(C(C)(C)C)cc2)cc1.[Br-].[Br-]', 'CC(C)(C)c1cc[n+](COC[n+]2ccc(C(C)(C)C)cc2)cc1.[Cl-].[Cl-]', 'CC(C)(C)c1cc[n+](CCOCC[n+]2ccc(C(C)(C)C)cc2)cc1.[Br-].[Br-]', 'CC(C)(C)c1cc[n+](C/C=C/C[n+]2ccc(C(C)(C)C)cc2)cc1.[Br-].[Br-]', 'CC(C)(C)c1cc[n+](C/C=C\\C[n+]2ccc(C(C)(C)C)cc2)cc1.[Cl-].[Cl-]', 'CC(C)(C)c1cc[n+](Cc2ccccc2C[n+]2ccc(C(C)(C)C)cc2)cc1.[Br-].[Br-]', 'CC(C)(C)c1cc[n+](Cc2cccc(C[n+]3ccc(C(C)(C)C)cc3)c2)cc1.[Br-].[Br-]', 'CC(C)(C)c1cc[n+](Cc2ccc(C[n+]3ccc(C(C)(C)C)cc3)cc2)cc1.[Br-].[Br-]', 'CC(C)(C)c1cc[n+](/C=C2\\C=CC3=CC/C(=C\\[n+]4ccc(C(C)(C)C)cc4)C=C3C2)cc1.[Br-].[Br-]', 'CC(C)(C)c1cc[n+](/C=C2\\C=CC3=CC/C(=C\\[n+]4ccc(C(C)(C)C)cc4)C=C3C2)cc1.[Br-].[Br-]', '[Br-].[Br-].c1ccc2c[n+](CCCCCCCCCC[n+]3ccc4ccccc4c3)ccc2c1', 'CN[C@H]1CCC2=CC3=CC[C@]4(C)[C@@H]([C@H](C)N(C)C)[C@H](O)C[C@@]4(C)[C@@H]3CC[C@H]2C1(C)C', 'CC(=O)O[C@@H]1C[C@@]2(C)[C@@H]3CC[C@H]4[C@](C)(CO)[C@@H](NC(=O)c5ccccc5)C=C[C@]4(O)CC3=CC[C@]2(C)[C@H]1[C@H](C)N(C)C', 'CC(=O)O[C@@H]1C[C@@]2(C)[C@@H]3C[C@H]4OC[C@@]5(C)[C@@H](NC(=O)c6ccccc6)C=C[C@@](O)(CC3=CC[C@]2(C)[C@H]1[C@H](C)N(C)C)[C@@H]45', 'CC(=O)O[C@H]1C[C@@H]2C(=CC[C@]3(C)[C@@H]([C@H](C)N(C)C)[C@H](O)C[C@@]23C)C[C@@]23O[C@@H]2[C@H]2OC[C@](C)([C@H]2NC(=O)c2ccccc2)[C@H]13', 'C[C@@H]([C@H]1[C@H](O)C[C@@]2(C)[C@@H]3C[C@H](O)[C@H]4[C@]5(C)CO[C@]4(C=C[C@@H]5NC(=O)c4ccccc4)CC3=CC[C@]12C)N(C)C', 'C[C@@H]([C@H]1[C@H](O)C[C@@]2(C)[C@@H]3C=C[C@H]4C(C)(C)[C@@H](N(C)C)[C@H](O)C[C@@]45C[C@@]35CC[C@]12C)N(C)C', 'C[C@@H]1[C@@H]2CC[C@@H]3C(=CC[C@]4(C)[C@@H]([C@H](C)N(C)C)[C@H](O)C[C@@]34C)C=C2CC[C@@H]1N(C)C', 'C[C@@H]([C@H]1[C@H](O)C[C@@]2(C)[C@@H]3CC[C@@H]4C(=CC3=CC[C@]12C)CC[C@H](N(C)C)C4(C)C)N(C)C', 'C[C@@H]([C@H]1CC[C@@]2(C)[C@@H]3CC[C@@H]4C(=CC3=CC[C@]12C)CC[C@H](N(C)C)C4(C)C)N(C)C', 'C/C=C1/C(=O)C[C@@]2(C)[C@@H]3CC[C@H]4C(C)(C)[C@@H](N(C)C)CC[C@@]45C[C@@]35CC[C@]12C', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)NC(c1ccccc1)N2', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)CNCCC2', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)CCNCC2', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)CNCCO2', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1cc2c3c(c1)CC(=O)N3CCC2', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1cc2c3c(c1)CCN3C(=O)CC2', 'O=C(CCC1CCN(Cc2ccccc2F)CC1)c1cc2c3c(ccn3C(=O)CC2)c1', 'O=C(CCC1CCN(Cc2cccc(F)c2)CC1)c1cc2c3c(ccn3C(=O)CC2)c1', 'O=C(CCC1CCN(Cc2ccccc2Cl)CC1)c1cc2c3c(ccn3C(=O)CC2)c1', 'O=C(CCC1CCN(Cc2cccc(Cl)c2)CC1)c1cc2c3c(ccn3C(=O)CC2)c1', 'O=C(CCC1CCN(Cc2ccc(Cl)cc2)CC1)c1cc2c3c(ccn3C(=O)CC2)c1', 'COc1ccccc1CN1CCC(CCC(=O)c2cc3c4c(ccn4C(=O)CC3)c2)CC1', 'COc1cccc(CN2CCC(CCC(=O)c3cc4c5c(ccn5C(=O)CC4)c3)CC2)c1', 'COc1ccc(CN2CCC(CCC(=O)c3cc4c5c(ccn5C(=O)CC4)c3)CC2)cc1', 'O=C(CCC1CCN(Cc2ccccc2O)CC1)c1cc2c3c(ccn3C(=O)CC2)c1', 'O=C(CCC1CCN(Cc2cccc(O)c2)CC1)c1cc2c3c(ccn3C(=O)CC2)c1', 'O=C(CCC1CCN(Cc2ccccc2[N+](=O)[O-])CC1)c1cc2c3c(ccn3C(=O)CC2)c1', 'O=C(CCC1CCN(Cc2cccc([N+](=O)[O-])c2)CC1)c1cc2c3c(ccn3C(=O)CC2)c1', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1cc2c3c(c1)CCC(=O)N3CCC2', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1cc2c3c(c1)CCN3C(=O)CCC2', 'CS(=O)(=O)Nc1ccc2c(CCC3CCN(Cc4ccccc4)CC3)noc2c1', 'c1ccc(CN2CCC(CCc3noc4cc(C5CCCCC5)ccc34)CC2)cc1', 'O=C(CCC1CCN(Cc2ccc(F)cc2)CC1)c1cc2c3c(ccn3C(=O)CC2)c1', 'O=C(CCC1CCN(Cc2ccc(O)cc2)CC1)c1cc2c3c(ccn3C(=O)CC2)c1', 'O=C(CCC1CCN(Cc2ccc([N+](=O)[O-])cc2)CC1)c1cc2c3c(ccn3C(=O)CC2)c1', 'Cc1ccc([C@@H](C)CC(=O)CC(C)C)cc1', 'C[N+](C)(CCNC(=O)C(=O)NCC[N+](C)(C)Cc1ccccc1Cl)Cc1ccccc1Cl.[Cl-].[Cl-]', 'C[N+](C)(CCNC(=O)C(=O)NCC[N+](C)(C)Cc1ccccc1Cl)Cc1ccccc1Cl.[Cl-].[Cl-]', 'C[N+](C)(CCNC(=O)C(=O)NCC[N+](C)(C)Cc1ccccc1Cl)Cc1ccccc1Cl.[Cl-].[Cl-]', 'C[N+](C)(CCNC(=O)C(=O)NCC[N+](C)(C)Cc1ccccc1Cl)Cc1ccccc1Cl.[Cl-].[Cl-]', 'C[N+](C)(CCNC(=O)C(=O)NCC[N+](C)(C)Cc1ccccc1Cl)Cc1ccccc1Cl.[Cl-].[Cl-]', 'C[N+](C)(CCNC(=O)C(=O)NCC[N+](C)(C)Cc1ccccc1Cl)Cc1ccccc1Cl.[Cl-].[Cl-]', 'C[N+](C)(CCNC(=O)C(=O)NCC[N+](C)(C)Cc1ccccc1Cl)Cc1ccccc1Cl.[Cl-].[Cl-]', 'C[N+](C)(CCNC(=O)C(=O)NCC[N+](C)(C)Cc1ccccc1Cl)Cc1ccccc1Cl.[Cl-].[Cl-]', 'C[N+](C)(CCNC(=O)C(=O)NCC[N+](C)(C)Cc1ccccc1Cl)Cc1ccccc1Cl.[Cl-].[Cl-]', '[Br-].[Br-].c1ccc2c[n+](CCCC[n+]3ccc4ccccc4c3)ccc2c1', '[Br-].[Br-].c1ccc2c[n+](CCCCC[n+]3ccc4ccccc4c3)ccc2c1', '[Br-].[Br-].c1ccc2c[n+](CCCCCC[n+]3ccc4ccccc4c3)ccc2c1', '[Br-].[Br-].c1ccc2c[n+](CCCCCCC[n+]3ccc4ccccc4c3)ccc2c1', '[Br-].[Br-].c1ccc2c[n+](CCCCCCCC[n+]3ccc4ccccc4c3)ccc2c1', '[Br-].[Br-].c1ccc2c[n+](CCCCCCCC[n+]3ccc4ccccc4c3)ccc2c1', '[Br-].[Br-].c1ccc2c[n+](CCCCCCCC[n+]3ccc4ccccc4c3)ccc2c1', '[Cl-].[Cl-].c1ccc2c[n+](COC[n+]3ccc4ccccc4c3)ccc2c1', 'C(=C/C[n+]1ccc2ccccc2c1)\\C[n+]1ccc2ccccc2c1.[Br-].[Br-]', 'C(=C\\C[n+]1ccc2ccccc2c1)\\C[n+]1ccc2ccccc2c1.[Cl-].[Cl-]', '[Br-].[Br-].c1cc(C[n+]2ccc3ccccc3c2)cc(C[n+]2ccc3ccccc3c2)c1', '[Br-].[Br-].c1ccc2c[n+](Cc3ccc(C[n+]4ccc5ccccc5c4)cc3)ccc2c1', '[Br-].[Br-].c1ccc2c[n+](Cc3ccc4ccc(C[n+]5ccc6ccccc6c5)cc4c3)ccc2c1', '[Br-].[Br-].c1ccc2c[n+](Cc3ccc4ccc(C[n+]5ccc6ccccc6c5)cc4c3)ccc2c1', '[Br-].[Br-].c1ccc2c[n+](Cc3ccc4ccc(C[n+]5ccc6ccccc6c5)cc4c3)ccc2c1', 'COc1ccc(C(=O)NC2CCN(Cc3c(F)cccc3Cl)CC2)cc1OC', 'c1ccc(CN2CCc3[nH]c4ccccc4c3C2)cc1', 'CC(C)(O)CCc1c(-c2ccc(O)cc2O)oc2c3c(cc(O)c2c1=O)OC(C)(C)C=C3', 'O=C(NC1CCN(Cc2ccccc2)CC1)c1ccc2c(c1)OCCCO2', '[Br-].[Br-].c1ccc2c(c1)ccc[n+]2CCC[n+]1cccc2ccccc21', '[Br-].[Br-].c1ccc2c(c1)ccc[n+]2CCCC[n+]1cccc2ccccc21', '[Br-].[Br-].c1ccc2c(c1)ccc[n+]2CCCCC[n+]1cccc2ccccc21', '[Br-].[Br-].c1ccc2c(c1)ccc[n+]2CCCCCCC[n+]1cccc2ccccc21', '[Br-].[Br-].c1ccc2c(c1)ccc[n+]2CCCCCCCC[n+]1cccc2ccccc21', '[Br-].[Br-].c1ccc2c(c1)ccc[n+]2CCOCC[n+]1cccc2ccccc21', 'C(=C/C[n+]1cccc2ccccc21)\\C[n+]1cccc2ccccc21.[Br-].[Br-]', 'C(=C\\C[n+]1cccc2ccccc21)\\C[n+]1cccc2ccccc21.[Cl-].[Cl-]', '[Br-].[Br-].c1ccc(C[n+]2cccc3ccccc32)c(C[n+]2cccc3ccccc32)c1', '[Br-].[Br-].c1cc(C[n+]2cccc3ccccc32)cc(C[n+]2cccc3ccccc32)c1', '[Br-].[Br-].c1ccc2c(c1)ccc[n+]2Cc1ccc2ccc(C[n+]3cccc4ccccc43)cc2c1', '[Br-].[Br-].c1ccc2c(c1)ccc[n+]2Cc1ccc2ccc(C[n+]3cccc4ccccc43)cc2c1', '[Br-].[Br-].c1ccc2c(c1)ccc[n+]2Cc1ccc2ccc(C[n+]3cccc4ccccc43)cc2c1', 'COc1ccc(C2c3c(nc4c(c3N)CCCC4)Oc3ccc4ccccc4c32)cc1OC', 'CC(=O)N1CCN(c2nccc(NCc3ccccc3)n2)CC1', 'CC(C)(C)OC(=O)N1CCN(c2nccc(NCc3ccccc3)n2)CC1', 'c1ccc(CNc2ccnc(N3CCN(C4CCCCC4)CC3)n2)cc1', 'CC(C)N1CCN(c2nccc(NCc3ccccc3)n2)CC1', 'CC(C)C1CCN(c2nccc(NCc3ccccc3)n2)CC1', 'CCCN1CCN(c2nccc(NCc3ccccc3)n2)CC1', 'OCCN1CCN(c2nccc(NCc3ccccc3)n2)CC1', 'COCCN1CCN(c2nccc(NCc3ccccc3)n2)CC1', 'Clc1ccc(CN2CCN(c3nccc(NCc4ccccc4)n3)CC2)cc1', 'Brc1ccc(CN2CCN(c3nccc(NCc4ccccc4)n3)CC2)cc1', 'Fc1ccc(CN2CCN(c3nccc(NCc4ccccc4)n3)CC2)cc1', 'FC(F)(F)c1ccc(CN2CCN(c3nccc(NCc4ccccc4)n3)CC2)cc1', 'c1ccc(CNc2ccnc(NC3CCN(Cc4ccccc4)CC3)n2)cc1', '[O-][S+]1CCN(c2nccc(NCc3ccccc3)n2)CC1', 'O=S1(=O)CCN(c2nccc(NCc3ccccc3)n2)CC1', 'c1ccc(CNc2ccnc(N3CCNCC3)n2)cc1', 'CC(=O)N1CCN(c2nccc(NCCc3ccccc3)n2)CC1', 'CC1CCN(c2nccc(NCc3cccc4ccccc34)n2)CC1', 'CC1CCN(c2nccc(NCc3cccc4ccccc34)n2)CC1', 'CC(C)=CC1Oc2cc(O)ccc2-c2oc3c4c(cc(O)c3c(=O)c21)OC(C)(C)C=C4', 'CC1(C)C=Cc2c(cc(O)c3c(=O)c4c(oc23)-c2ccc(O)cc2OC(C(C)(C)O)C4)O1', 'COc1cc(CNCCCNC2=CC(=O)C(NCCCNCc3ccc(O)c(OC)c3)=CC2=O)ccc1O', 'COc1cc(CNCCCCCCNC2=CC(=O)C(NCCCCCCNCc3ccc(O)c(OC)c3)=CC2=O)ccc1O', 'O=C1C=C(NCCCNCc2cc3ccccc3o2)C(=O)C=C1NCCCNCc1cc2ccccc2o1', 'O=C1C=C(NCCCCCCNCc2nc3ccccc3s2)C(=O)C=C1NCCCCCCNCc1nc2ccccc2s1', 'O=C1C=C(NCCCNCc2ccc(-c3nc4ccccc4s3)cc2)C(=O)C=C1NCCCNCc1ccc(-c2nc3ccccc3s2)cc1', 'O=C1C=C(NCCCCCCNCc2ccc(-c3nc4ccccc4s3)cc2)C(=O)C=C1NCCCCCCNCc1ccc(-c2nc3ccccc3s2)cc1', '[Br-].[Br-].c1cc(C[N+]23CCC(CC2)CC3)cc(-c2cccc(C[N+]34CCC(CC3)CC4)c2)c1', 'COc1cc2c(cc1OC)CN(c1cc[n+](Cc3ccc(C[n+]4ccc(N5CCCCC5)cc4)cc3)cc1)CC2.[Br-].[Br-]', 'CCN(CC)CC(=O)Nc1ccc(C(=O)c2ccc(CN(C)Cc3ccccc3)cc2)cc1', 'CN(Cc1ccccc1)Cc1ccc(C(=O)c2ccc(NC(=O)CN3CCCCC3)cc2)cc1', 'CN(Cc1ccccc1)Cc1ccc(C(=O)c2ccc(NC(=O)CN3CCOCC3)cc2)cc1', 'CN(Cc1ccccc1)Cc1ccc(C(=O)c2ccc(NC(=O)CCCCN3CCCCC3)cc2)cc1', 'CCN(CC)CCCOc1ccc(C(=O)c2ccc(CN(C)Cc3ccccc3)cc2)cc1', 'CN(Cc1ccccc1)Cc1ccc(C(=O)c2ccc(OCCCN3CCCCC3)cc2)cc1', 'CN(Cc1ccccc1)Cc1ccc(C(=O)c2ccc(OCCCN3CCOCC3)cc2)cc1', 'CN(Cc1ccccc1)Cc1ccc(C(=O)c2ccc(OCCCN3CCC(n4c(=O)[nH]c5ccccc54)CC3)cc2)cc1', 'CCN(CC)CCCCCOc1ccc(C(=O)c2ccc(CN(C)Cc3ccccc3)cc2)cc1', 'CN(Cc1ccccc1)Cc1ccc(C(=O)c2ccc(OCCCCCN3CCCCC3)cc2)cc1', 'CCN(CC)CCCCCCOc1ccc(C(=O)c2ccc(CN(C)Cc3ccccc3)cc2)cc1', 'CN(Cc1ccccc1)Cc1ccc(C(=O)c2ccc(OCCCCCCN3CCCCC3)cc2)cc1', 'CCN(CC)CCCCCCCOc1ccc(C(=O)c2ccc(CN(C)Cc3ccccc3)cc2)cc1', 'CN(Cc1ccccc1)Cc1ccc(C(=O)c2ccc(OCCCCCCCN3CCCCC3)cc2)cc1', 'CCN(CC)CCc1ccc(Oc2ccc(C(=O)c3ccc(CN(C)Cc4ccccc4)cc3)cc2)cc1', 'CN(C)CC(C)(C)c1cccc(OC(=O)NCCCCN2CCOCC2)c1', 'CN(C)CC(C)(C)c1cccc(OC(=O)NCCCCN2CCOCC2)c1', 'CN(C)CC(C)(C)c1cccc(OC(=O)NCCCCN2CCOCC2)c1', 'CN(C)CC(C)(C)c1cccc(OC(=O)NCCCCN2CCOCC2)c1', 'CCN(CC)C(C)C(=O)N1c2ccccc2Sc2ccccc21.Cl', 'CCN(CC)C(C)C(=O)N1c2ccccc2Sc2ccccc21.Cl', 'CC(C)C1=N[C@H]2CC[C@]34C[C@]35C(=O)C[C@]3(C)[C@@H]([C@H](C)N(C)CCCN6C(=O)c7ccccc7C6=O)[C@H](O)C[C@@]3(C)[C@@H]5CC[C@H]4[C@]2(C)CS1', 'COc1cccc(CN(C)[C@@H](C)[C@H]2[C@H](O)C[C@@]3(C)[C@@H]4CC[C@H]5[C@]6(C)CSC(C(C)C)=N[C@H]6CC[C@@]56C[C@@]46C(=O)C[C@]23C)c1', 'COc1ccc(CN(C)[C@@H](C)[C@H]2[C@H](O)C[C@@]3(C)[C@@H]4CC[C@H]5[C@]6(C)CSC(C(C)C)=N[C@H]6CC[C@@]56C[C@@]46C(=O)C[C@]23C)cc1', 'CC(C)C1=N[C@H]2CC[C@]34C[C@]35C(=O)C[C@]3(C)[C@@H]([C@H](C)N(C)Cc6ccccc6C(F)(F)F)[C@H](O)C[C@@]3(C)[C@@H]5CC[C@H]4[C@]2(C)CS1', 'CC(C)C1=N[C@H]2CC[C@]34C[C@]35C(=O)C[C@]3(C)[C@@H]([C@H](C)N(C)Cc6cccc(C(F)(F)F)c6)[C@H](O)C[C@@]3(C)[C@@H]5CC[C@H]4[C@]2(C)CS1', 'CC(C)C1=N[C@H]2CC[C@]34C[C@]35C(=O)C[C@]3(C)[C@@H]([C@H](C)N(C)Cc6ccc(C(F)(F)F)cc6)[C@H](O)C[C@@]3(C)[C@@H]5CC[C@H]4[C@]2(C)CS1', 'CC(C)C1=N[C@H]2CC[C@]34C[C@]35C(=O)C[C@]3(C)[C@@H]([C@H](C)N(C)CCN6C(=O)c7ccccc7C6=O)[C@H](O)C[C@@]3(C)[C@@H]5CC[C@H]4[C@]2(C)CS1', 'CC(C)C1=N[C@H]2CC[C@]34C[C@]35C(=O)C[C@]3(C)[C@@H]([C@H](C)N(C)C)[C@H](O)C[C@@]3(C)[C@@H]5CC[C@H]4[C@]2(C)CS1', 'CN[C@@H](C)[C@H]1[C@H](O)C[C@@]2(C)[C@@H]3CC[C@H]4[C@]5(C)CSC(C(C)C)=N[C@H]5CC[C@@]45C[C@@]35C(=O)C[C@]12C', 'CC(=O)N(C)[C@@H](C)[C@H]1[C@H](O)C[C@@]2(C)[C@@H]3CC[C@H]4[C@]5(C)CSC(C(C)C)=N[C@H]5CC[C@@]45C[C@@]35C(=O)C[C@]12C', 'CC(C)C1=N[C@H]2CC[C@]34C[C@]35C(=O)C[C@]3(C)[C@@H]([C@H](C)N(C)C(=O)c6ccccc6)[C@H](O)C[C@@]3(C)[C@@H]5CC[C@H]4[C@]2(C)CS1', 'CC(C)C1=N[C@H]2CC[C@]34C[C@]35C(=O)C[C@]3(C)[C@@H]([C@H](C)N(C)C(=O)OC(C)(C)C)[C@H](O)C[C@@]3(C)[C@@H]5CC[C@H]4[C@]2(C)CS1', 'CC(C)C1=N[C@H]2CC[C@]34C[C@]35C(=O)C[C@]3(C)[C@@H]([C@H](C)N(C)Cc6ccccc6)[C@H](O)C[C@@]3(C)[C@@H]5CC[C@H]4[C@]2(C)CS1', 'CC(C)C1=N[C@H]2CC[C@]34C[C@]35C(=O)C[C@]3(C)[C@@H]([C@H](C)N(C)CCc6ccccc6)[C@H](O)C[C@@]3(C)[C@@H]5CC[C@H]4[C@]2(C)CS1', 'CC(C)C1=N[C@H]2CC[C@]34C[C@]35C(=O)C[C@]3(C)[C@@H]([C@H](C)N(C)CCCc6ccccc6)[C@H](O)C[C@@]3(C)[C@@H]5CC[C@H]4[C@]2(C)CS1', 'COc1ccccc1CN(C)[C@@H](C)[C@H]1[C@H](O)C[C@@]2(C)[C@@H]3CC[C@H]4[C@]5(C)CSC(C(C)C)=N[C@H]5CC[C@@]45C[C@@]35C(=O)C[C@]12C', 'C[C@@H]([C@H]1[C@H](O)C[C@@]2(C)[C@@H]3CC[C@H]4[C@]5(C)COC(C(C)(C)F)=N[C@H]5CC[C@@]45C[C@@]35C(=O)C[C@]12C)N(C)C', 'C[C@@H]([C@H]1[C@H](O)C[C@@]2(C)[C@@H]3CC[C@H]4[C@]5(C)COC(C(F)(C(F)(F)F)C(F)(F)F)=N[C@H]5CC[C@@]45C[C@@]35C(=O)C[C@]12C)N(C)C', 'CC(C)C1=N[C@H]2CC[C@]34C[C@]35C(=O)C[C@]3(C)[C@@H]([C@H](C)N(C)C)[C@H](O)C[C@@]3(C)[C@@H]5CC[C@H]4[C@]2(C)C=C1', 'CC(C)C1=N[C@H]2CC[C@]34C[C@]35C(=O)C[C@]3(C)[C@@H]([C@H](C)N(C)C)[C@H](O)C[C@@]3(C)[C@@H]5CC[C@H]4[C@]2(C)CC1', 'C[C@@H](c1cccc(OC(=O)N(C)[C@@H](C)Cc2ccccc2)c1)N(C)C.Cl', 'C[C@H](Cc1ccccc1)NC(=O)Oc1cccc([C@H](C)N(C)C)c1.Cl', 'C[C@@H](Cc1ccccc1)NC(=O)Oc1cccc([C@H](C)N(C)C)c1.Cl', 'C[C@H](Cc1ccccc1)N(C)C(=O)Oc1cccc([C@H](C)N(C)C)c1.Cl', 'C#CCN(C)[C@H](C)Cc1cccc(OC(=O)N(C)C2CCCCC2)c1', 'C#CCN(C)[C@H](C)Cc1cccc(OC(=O)N(C)CCCCCC)c1', 'CCN(C)C(=O)Oc1ccc2c(c1)[C@H](N)CC2.Cl', 'C#CCN[C@@H]1CCc2ccc(OC(=O)N(C)C)cc21.Cl', 'C#CCN[C@@H]1CCc2ccc(OC(=O)N(C)CC)cc21.Cl', 'C#CCN[C@H]1CCc2ccc(OC(=O)N(C)CCC)cc21.Cl', 'C#CCN[C@@H]1CCc2ccc(OC(=O)N(C)CCCC)cc21.Cl', 'CNC(=C[N+](=O)[O-])NCCSCc1ccc(CN(C)C)o1', 'CNC(=C[N+](=O)[O-])NCCSCc1ccc(CN(C)C)o1', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(CCCCCCCCN1C(=O)c3ccccc3S1(=O)=O)C2.O=C(O)/C=C/C(=O)O', 'O=C(c1ccccc1)N1c2ccccc2Sc2ccc(Cl)cc21', 'COc1ccc(C(=O)N2c3ccccc3Sc3ccc(Cl)cc32)cc1', 'Cl.O=C(NCCNc1c2c(nc3ccccc13)CCCC2)c1ccc(F)cc1', 'Cl.O=C(NCCCNc1c2c(nc3ccccc13)CCCC2)c1ccc(F)cc1', 'Cl.O=C(NCCCCNc1c2c(nc3ccccc13)CCCC2)c1ccc(F)cc1', 'c1ccc2c(NCCCCCCN3CCOCC3)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCCCCCCN3CCOCC3)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCCCCCCCN3CCOCC3)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCCCCCCCCN3CCOCC3)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCCCCCCCCCN3CCOCC3)c3c(nc2c1)CCCC3', 'O=c1c2ccc(OCCCCCCCNc3c4c(nc5ccccc35)CCCC4)cc2oc2ncccc12', 'c1ccc2c(c1)Cc1c-2nc2ccccc2c1NCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'Clc1ccc2c(NCCCCCCCNc3c4c(nc5ccccc35)-c3ccccc3C4)c3c(nc2c1)CCCC3', 'Clc1ccc2c(NCCCCCCCNc3c4c(nc5ccccc35)-c3ccccc3C4)c3c(nc2c1)CCCC3', 'Clc1ccc2c(NCCCCCCCNc3c4c(nc5cc(Cl)ccc35)-c3ccccc3C4)c3c(nc2c1)CCCC3', 'Clc1ccc2c(NCCCCCCCOc3ccc(-c4cc5ccccc5o4)cc3)c3c(nc2c1)CCCC3', 'Clc1ccc2c(NCCCCCCCOc3cccc(-c4cc5ccccc5o4)c3)c3c(nc2c1)CCCC3', 'c1ccc2c(c1)Cc1c-2nc2ccccc2c1NCCCCCCCNc1c2c(nc3ccccc13)-c1ccccc1C2', 'CCC(c1ccc(O)c(OC)c1)c1cccc(O)c1O', 'CC(C)N1CCN(c2nccc(NCc3cccc4ccccc34)n2)CC1', 'CC(C)C1CCN(c2nccc(NCc3cccc4ccccc34)n2)CC1', 'CCCN1CCN(c2nccc(NCc3cccc4ccccc34)n2)CC1', 'c1ccc2c(CNc3ccnc(N4CCN(C5CCCCC5)CC4)n3)cccc2c1', 'CC(=O)N1CCN(c2nccc(NCc3cccc4ccccc34)n2)CC1', 'CC(C)(C)OC(=O)N1CCN(c2nccc(NCc3cccc4ccccc34)n2)CC1', 'c1ccc2c(CNc3ccnc(N4CCNCC4)n3)cccc2c1', 'OCCN1CCN(c2nccc(NCc3cccc4ccccc34)n2)CC1', 'COCCN1CCN(c2nccc(NCc3cccc4ccccc34)n2)CC1', 'CN1CCN(c2nccc(NC(c3ccccc3)c3ccccc3)n2)CC1', 'CC1CCN(c2nccc(NC(c3ccccc3)c3ccccc3)n2)CC1', 'CC(C)N1CCN(c2nccc(NC(c3ccccc3)c3ccccc3)n2)CC1', 'CC(C)C1CCN(c2nccc(NC(c3ccccc3)c3ccccc3)n2)CC1', 'CCCN1CCN(c2nccc(NC(c3ccccc3)c3ccccc3)n2)CC1', 'c1ccc(C(Nc2ccnc(N3CCN(C4CCCCC4)CC3)n2)c2ccccc2)cc1', 'CC(=O)N1CCN(c2nccc(NC(c3ccccc3)c3ccccc3)n2)CC1', 'c1ccc(C(Nc2ccnc(N3CCNCC3)n2)c2ccccc2)cc1', 'OCCN1CCN(c2nccc(NC(c3ccccc3)c3ccccc3)n2)CC1', 'COCCN1CCN(c2nccc(NC(c3ccccc3)c3ccccc3)n2)CC1', 'CCCCC/C=C\\C/C=C\\CCCCCCCC(=O)OCCCc1cc(OC)c2oc(-c3ccc4c(c3)OCO4)cc2c1', 'CCCCC/C=C\\C/C=C\\CCCCCCCC(=O)OCCCc1ccc2oc(-c3ccc4c(c3)OCO4)cc2c1', 'CCCCCCCC/C=C\\CCCCCCCC(=O)OCCCc1cc(OC)c2oc(-c3ccc4c(c3)OCO4)cc2c1', 'CCCCCCCC/C=C\\CCCCCCCC(=O)OCCCc1ccc2oc(-c3ccc4c(c3)OCO4)cc2c1', 'CC[C@H](C)C(=O)OCCCc1cc(OC)c2oc(-c3ccc4c(c3)OCO4)cc2c1', 'CC[C@H](C)C(=O)OCCCc1ccc2oc(-c3ccc4c(c3)OCO4)cc2c1', 'COc1cc(CCCOC(=O)Nc2ccccc2)cc2cc(-c3ccc4c(c3)OCO4)oc12', 'O=c1c2ccccc2nc2n1CCC2O', 'O=C(CCCC[C@H]1CCSS1)NCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCCC[C@@H]1CCSS1)NCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCCCc1cccs1)NCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'CCN(CCCCCCNC(=O)CCCCC1CCSS1)Cc1ccccc1OC', 'COc1cccc(C(C)N(C)CCCNC(=O)CCCCC2CCSS2)c1', 'CC(c1cccc(OCCCNC(=O)CCCCC2CCSS2)c1)N(C)C', 'C=CC(CBr)Nc1c2c(nc3ccc(OC)cc13)CCCC2', 'COC(=O)[C@H](C)CC(=O)C[C@@H](C)[C@H]1C[C@@H]2OC(C)(C)O[C@H]3C[C@H]4C(C)(C)C(=O)CC[C@]4(C)C4=C3[C@@]2(C)[C@]1(C)CC4=O', 'CCCCOC(=O)C(C)CC(=O)C[C@@H](C)[C@H]1CC(=O)[C@@]2(C)C3=C(C(=O)[C@@H](OC(C)=O)[C@]12C)[C@@]1(C)CC[C@H](O)C(C)(C)[C@@H]1CC3=O', 'C/C(=C\\CC[C@@H](C)[C@H]1CC[C@@]2(C)C3=CC[C@H]4C(C)(C)[C@@H](O)CC[C@]4(C)C3=CC[C@]12C)C(=O)O', 'C/C(=C\\CC[C@@H](C)[C@H]1CC[C@@]2(C)C3=CC[C@H]4C(C)(C)[C@@H](O)CC[C@]4(C)C3=CC[C@]12C)C(=O)O', 'C[C@H](CCC=C(CO)CO)[C@H]1CC[C@@]2(C)C3=CC[C@H]4C(C)(C)C(=O)CC[C@]4(C)C3=CC[C@]12C', 'C[C@H](CC[C@H](O)C(C)(C)O)[C@H]1CC[C@@]2(C)C3=CC[C@H]4C(C)(C)[C@@H](O)CC[C@]4(C)C3=CC[C@]12C', 'C[C@H](CC[C@H](O)[C@](C)(O)CO)[C@H]1CC[C@@]2(C)C3=CC[C@H]4C(C)(C)C(=O)CC[C@]4(C)C3=CC[C@]12C', 'C/C(=C\\CC[C@@H](C)[C@H]1CC[C@@]2(C)C3=C(CC[C@]12C)[C@@]1(C)CC[C@H](O)C(C)(C)[C@@H]1CC3=O)CO', 'CCCCOC(=O)CC[C@@H](C)[C@H]1CC[C@@]2(C)C3=C(C(=O)C[C@]12C)[C@@]1(C)CC[C@H](O)C(C)(C)[C@@H]1C[C@@H]3O', 'CCCCOC(=O)CC[C@@H](C)[C@H]1CC[C@@]2(C)C3=C(C(=O)C[C@]12C)[C@@]1(C)CCC(=O)C(C)(C)[C@@H]1C[C@@H]3O', 'CCCOCCn1c(=O)nc(NC2CCC(O)CC2)c2nnc(-c3ccc(OC)nc3)cc21', 'Nc1c2c(nc3c1cc(-c1ccccc1)n3-c1ccccc1)CCCC2', 'C=CCCCNc1c2c(nc3ccc(OC)cc13)CCCC2', 'COc1ccc2nc3c(c(N4CCCCCN(c5c6c(nc7ccc(OC)cc57)CCCC6)CCCCC4)c2c1)CCCC3', 'C#CCOP(C)(=O)OC(C(F)(F)F)C(F)(F)F', 'C#CCOP(C)(=O)OC(C(=O)OC)C(F)(F)F', 'COc1c(CC=C(C)C)c(O)cc2c1C[C@H](c1ccc(O)c(CC=C(C)C)c1O)CO2', 'C#CCN(C)Cc1cc2cc(OCCCC3CCN(Cc4ccccc4)CC3)ccc2n1C', 'C#CCN(C)Cc1cc2cc(OCCCC3CCN(Cc4ccccc4)CC3)ccc2n1C', 'C#CCN(C)Cc1cc2cc(OCCCC3CCN(Cc4ccccc4)CC3)ccc2n1C', 'C#CCN(C)Cc1cc2cc(OCCCC3CCN(Cc4ccccc4)CC3)ccc2n1C', 'C#CCN(C)Cc1cc2cc(OCCCC3CCN(Cc4ccccc4)CC3)ccc2n1C', 'C#CCN(C)Cc1cc2cc(OCCCC3CCN(Cc4ccccc4)CC3)ccc2n1C', 'CCN(CCCCCCNC1=CC(=O)c2ccccc2C1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)c2ccccc2C1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)c2ncccc2C1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(C)=C(C)C1=O)Cc1ccccc1OC', 'CCN(CCCCCCNC1=CC(=O)C(OC)=CC1=O)Cc1ccccc1OC', 'COc1ccccc1CNCCCCCCNC1=CC(=O)c2ccccc2C1=O', 'NCCCCCCNC1=CC(=O)c2ccccc2C1=O', 'COc1ccccc1CN1CCC(N2CCC(NC3=CC(=O)c4ccccc4C3=O)CC2)CC1', 'COC1=CC(=O)C(N2CCC(C3CCN(Cc4ccccc4OC)CC3)CC2)=CC1=O', 'COc1ccccc1NCc1cc(=O)oc2cc3oc4ccccc4c3cc12', 'COc1cc2c(C)c(OC)c(=O)oc2cc1OCCCCOc1cc(O)cc(N(C)C)c1', 'COc1cc2oc(=O)c(OCCCCOc3cc(O)cc(N(C)C)c3)c(C)c2cc1OC', 'CCCC/N=C(/C=N/O)N(C)C.Cl', 'CCCC/N=C(/C=N/O)N(C)C.Cl', 'CCCCC/N=C(/C=N/O)N(C)C.Cl', 'CC(C)CC/N=C(/C=N/O)N(C)C.Cl', 'Cl.O/N=C/C1=NCCN1Cc1ccccc1', 'Cl.O/N=C/C1=NCCN1Cc1ccc(Cl)cc1Cl', 'Cl.O/N=C/C1=NCCN1Cc1cc(C(F)(F)F)cc(C(F)(F)F)c1', 'CCCCCN1CCCN=C1/C=N/O.Cl', 'Cl.O/N=C/C1=NCCCN1Cc1ccccc1', 'Cl.O/N=C/C1=NCCCN1Cc1ccc(Cl)cc1Cl', 'Cl.O/N=C/C1=NCCCN1Cc1cc(C(F)(F)F)cc(C(F)(F)F)c1', 'CCCCN1CC(C)(C)CN=C1/C=N/O.Cl', 'CCCCCN1CC(C)(C)CN=C1/C=N/O.Cl', 'O=[N+]([O-])c1ccc2c(c1)c(OCc1ccccc1)nn2Cc1ccccc1', 'COc1cc([C@H]2OC[C@H]3[C@@H]2CO[C@@H]3c2cc(OC)c3c(c2)OCO3)cc(OC)c1OC', 'CC1=C[C@@H]2Cc3nc4cc(Cl)ccc4c(N)c3[C@H](C1)C2.Cl', 'Cc1ccc(S(=O)(=O)SCc2ccccc2)cc1', 'Cc1ccc(S(=O)(=O)SCc2ccccc2)cc1', 'CC1=C[C@@H]2Cc3nc4cc(Cl)ccc4c(N)c3[C@H](C1)C2', 'CC1=C[C@@H]2Cc3nc4cc(Cl)ccc4c(N)c3[C@H](C1)C2', 'COc1ccc([C@H]2Oc3cc(O)c(C/C=C(\\C)CCC=C(C)C)c(O)c3C(=O)[C@@H]2O)cc1O', 'COc1ccc([C@H]2Oc3cc(O)c(C/C=C(\\C)CCC=C(C)C)c(O)c3C(=O)[C@@H]2O)cc1O', 'COc1c(O)cc(C2Oc3cc(O)c(C/C=C(\\C)CCC=C(C)C)c(O)c3C(=O)[C@@H]2O)cc1O', 'COc1c(O)cc(C2Oc3cc(O)c(C/C=C(\\C)CCC=C(C)C)c(O)c3C(=O)[C@@H]2O)cc1O', 'COc1ccc([C@@H]2CC(=O)c3c(cc(O)c(C/C=C(\\C)CCC=C(C)C)c3O)O2)cc1O', 'COc1ccc([C@@H]2CC(=O)c3c(cc(O)c(C/C=C(\\C)CCC=C(C)C)c3O)O2)cc1O', 'COc1c(O)cc(C2CC(=O)c3c(cc(O)c(C/C=C(\\C)CCC=C(C)C)c3O)O2)cc1O', 'COc1c(O)cc(C2CC(=O)c3c(cc(O)c(C/C=C(\\C)CCC=C(C)C)c3O)O2)cc1O', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCNc4c5c(nc6ccccc46)CCCC5)c3C(C1)C2.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCN(C)CCCNc4c5c(nc6ccccc46)CCCC5)c3C(C1)C2.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCNc4c5c(nc6ccccc46)CCCC5)c3C(C1)C2.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCNc4c5c(nc6ccccc46)CCCC5)c3C(C1)C2.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)c3C(C1)C2.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCN(C)CCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)c3C(C1)C2.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)c3C(C1)C2.Cl', 'CC1=C[C@H]2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2.Cl', 'CC1=C[C@H]2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2.Cl', 'COc1ccc2c(ccc[n+]2C)c1.[I-]', 'COc1ccc2c(ccc[n+]2Cc2ccccc2)c1.[I-]', 'CC[C@H]1C[C@@H]2CN3CCc4c([nH]c5cc([C@H]6C[C@H]7C(C(=O)OC)[C@H](Cc8c6[nH]c6ccccc86)N(C)C[C@H]7CC)ccc45)[C@](C(=O)OC)(C2)[C@H]13', 'Cl.O=C(NCCCCCCCCCCNc1c2c(nc3ccccc13)CCCC2)c1cc(=O)c2ccccc2o1', 'Cl.O=C(NCCCCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)c1cc(=O)c2ccccc2o1', 'COc1ccc2oc(C(=O)NCCCCCCCCCCNc3c4c(nc5ccccc35)CCCC4)cc(=O)c2c1.Cl', 'COc1ccc2oc(C(=O)NCCCCCCCCCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)cc(=O)c2c1.Cl', 'COc1cc(OC)c2c(=O)cc(C(=O)NCCCCCCCCCCNc3c4c(nc5ccccc35)CCCC4)oc2c1.Cl', 'COc1cc(OC)c2c(=O)cc(C(=O)NCCCCCCCCCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)oc2c1.Cl', 'COc1cc2oc(C(=O)NCCCCCCCCCCNc3c4c(nc5ccccc35)CCCC4)cc(=O)c2cc1OC.Cl', 'COc1cc2oc(C(=O)NCCCCCCCCCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)cc(=O)c2cc1OC.Cl', 'Cl.O=C(NCCCCCCCCCCNc1c2c(nc3ccccc13)CCCC2)c1cc(=O)c2cc(O)ccc2o1', 'Cl.O=C(NCCCCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)c1cc(=O)c2cc(O)ccc2o1', 'COc1cc(O)c2c(=O)cc(C(=O)NCCCCCCCCCCNc3c4c(nc5ccccc35)CCCC4)oc2c1.Cl', 'Cl.O=C(NCCCCCCCCCCNc1c2c(nc3ccccc13)CCCC2)c1cc(=O)c2c(O)cc(O)cc2o1', 'COc1cc(O)c2c(=O)cc(C(=O)NCCCCCCCCCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)oc2c1.Cl', 'Cl.O=C(NCCCCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)c1cc(=O)c2c(O)cc(O)cc2o1', 'Cl.O=C(NCCCCCCCCCCNc1c2c(nc3ccccc13)CCCC2)c1cc(=O)c2cc(O)c(O)cc2o1', 'Cl.O=C(NCCCCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)c1cc(=O)c2cc(O)c(O)cc2o1', 'CC1=C[C@H]2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2', 'C#CCN1CCc2nc3nc(N4CCCC4)c(C#N)cc3c(N)c2C1', 'Clc1ccccc1CNc1ccnc(NC2CCN(Cc3ccccc3)CC2)n1', 'Clc1cccc(CNc2ccnc(NC3CCN(Cc4ccccc4)CC3)n2)c1', 'Clc1ccc(CNc2ccnc(NC3CCN(Cc4ccccc4)CC3)n2)cc1', 'Brc1ccc(CNc2ccnc(NC3CCN(Cc4ccccc4)CC3)n2)cc1', 'Fc1ccc(CNc2ccnc(NC3CCN(Cc4ccccc4)CC3)n2)cc1', 'Cc1ccc(CNc2ccnc(NC3CCN(Cc4ccccc4)CC3)n2)cc1', 'COc1ccc(CNc2ccnc(NC3CCN(Cc4ccccc4)CC3)n2)cc1', 'COc1ccc(CNc2ccnc(NC3CCN(Cc4ccccc4)CC3)n2)cc1OC', 'COc1cc(CNc2ccnc(NC3CCN(Cc4ccccc4)CC3)n2)cc(OC)c1OC', 'c1ccc(CN2CCC(Nc3nccc(NCc4ccc5c(c4)OCO5)n3)CC2)cc1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc(-c3cc4ccccc4oc3=O)cc2)c1', 'CN(CCCOc1ccc2ccc(=O)oc2c1)Cc1cccc(OC(=O)NCCCCCc2ccccc2)c1', 'CN(CCCOc1ccc2ccc(=O)oc2c1)Cc1cccc(OC(=O)NCCCCCCCN2CCOCC2)c1', 'CN(CCCOc1ccc2c(=O)c3cccnc3oc2c1)Cc1cccc(OC(=O)NCCCCCc2ccccc2)c1', 'CN(CCCOc1ccc2c(=O)c3cccnc3oc2c1)Cc1cccc(OC(=O)NCCCCCCCN2CCOCC2)c1', 'O=C(CCN1CCCCC1)N1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1', 'O=C(CCN1CCCCC1)N1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1', 'O=C(CCN1CCCCC1)N1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1', 'CN(CCC(=O)N1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1)C1CCCCC1', 'CN(CCC(=O)N1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1)C1CCCCC1', 'CN(CCC(=O)N1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1)C1CCCCC1', 'CCN(CCC(=O)N1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1)C1CCCCC1', 'CCN(CCC(=O)N1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1)C1CCCCC1', 'CCN(CCC(=O)N1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1)C1CCCCC1', 'CN(CCC(=O)N1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1)c1ccccc1', 'CN(CCC(=O)N1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1)c1ccccc1', 'CN(CCC(=O)N1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1)c1ccccc1', 'CN(CCCN1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1)C1CCCCC1', 'CN(CCCN1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1)C1CCCCC1', 'CN(CCCN1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1)C1CCCCC1', 'CCN(CCCN1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1)C1CCCCC1', 'CN(CCCN1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1)c1ccccc1', 'CN(CCCN1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1)c1ccccc1', 'CN(CCCN1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1)c1ccccc1', 'O=C(CCNCCNc1c2c(nc3ccccc13)CCCC2)N1CCN(CCNc2c3c(nc4ccccc24)CCCC3)CC1', 'CC(C)=CCC/C(C)=C/Cc1ccc(O)c2nc3cccc(O)c3nc12', 'CC(=O)Oc1cccc2nc3c(OC(C)=O)ccc(C/C=C(\\C)CCC=C(C)C)c3nc12', 'Oc1cccc2nc3c(O)cccc3nc12', 'N#Cc1c(N)nc(NC2CCN(Cc3ccccc3)CC2)c(C#N)c1-c1ccccc1', 'N#Cc1c(N)nc(NCCC2CCN(Cc3ccccc3)CC2)c(C#N)c1-c1ccccc1', 'N#Cc1c(N)nc(NCCCC2CCN(Cc3ccccc3)CC2)c(C#N)c1-c1ccccc1', 'N#Cc1c(N)nc(NCCCCC2CCN(Cc3ccccc3)CC2)c(C#N)c1-c1ccccc1', 'N#Cc1cc(C#N)c(NC2CCN(Cc3ccccc3)CC2)nc1N', 'N#Cc1cc(C#N)c(NCCC2CCN(Cc3ccccc3)CC2)nc1N', 'N#Cc1cc(C#N)c(NCCCC2CCN(Cc3ccccc3)CC2)nc1N', 'N#Cc1cc(C#N)c(NCCCCC2CCN(Cc3ccccc3)CC2)nc1N', 'C#CCN[C@@H]1CCc2ccc(OC(=O)N(C)CCC)cc21.Cl', 'C#CCN[C@H]1CCc2ccc(OC(=O)N(C)CC)cc21.Cl', 'C#CCN[C@H]1CCc2ccc(OC(=O)N(C)C)cc21.Cl', 'C#CCN(C)[C@@H](C)Cc1cccc(OC(=O)N(C)C2CCCCC2)c1', 'C#CCN(C)[C@@H](C)Cc1cccc(OC(=O)N(C)CCCCCC)c1', 'C#CCN[C@H]1CCc2ccc(OC(=O)N(C)CCCC)cc21.Cl', 'CC1=C[C@@H]2Cc3nc4cc(Cl)ccc4c(NCCCN(C)CCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)c3[C@H](C1)C2.Cl.Cl.Cl', 'CC1=C[C@@H]2Cc3nc4cc(Cl)ccc4c(NCCCN(C)CCCCNc4c5c(nc6ccccc46)CCCC5)c3[C@H](C1)C2.Cl.Cl.Cl', 'O=C(CCCCCCC(=O)NCCCCCCN1CC[C@H]2COc3ccccc3[C@H]21)NCCCCCCN1CC[C@H]2COc3ccccc3[C@H]21', 'CN(C)c1ccccc1CNCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'CN(C)c1ccccc1CNCCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'CN(C)c1ccccc1CNCCCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'CCCCCCCNC(=O)Oc1ccc2c(c1)CN1CCc3ccccc3C1N2C', 'CCCCCCCNC(=O)Oc1ccc2c(c1)CN1CCc3ccccc3C1N2C', 'CCCCCCCNC(=O)Oc1ccc2c(c1)CN1CCc3ccccc3C1N2C', 'CCCCCCCNC(=O)Oc1ccc2c(c1)CN1CCc3ccccc3C1N2C', 'CCCCCCCNC(=O)Oc1ccc2c(c1)CN1CCc3ccccc3C1N2C', 'CN(C)CCOc1ccc(C(=O)/C=C/c2ccccc2)cc1', 'CN(C)CCOc1ccc(C(=O)/C=C/c2ccccc2)cc1', 'CN(C)CCOc1ccc(C(=O)/C=C/c2ccccc2)cc1', 'O=C(NCCc1c[nH]c2ccccc12)c1ccc(OCCCCn2c3ccccc3c3ccccc32)cc1', 'CN(C)c1cc[n+](CCCCOc2ccc(C(=O)NCCc3c[nH]c4ccccc34)cc2)cc1.[Br-]', 'O=C(NCCc1c[nH]c2ccccc12)c1ccc(OCCCC[n+]2cccc3ccccc32)cc1.[Br-]', 'CC(=O)c1ccccc1NC(=O)c1ccc(OCCCC[n+]2cccc3ccccc32)cc1.[Br-]', 'CC(=O)c1ccccc1NC(=O)c1ccc(OCCCC[n+]2ccc3ccccc3c2)cc1.[Br-]', 'CC(=O)c1ccccc1NC(=O)c1ccc(OCCCCC[n+]2ccc3ccccc3c2)cc1.[Br-]', 'CC(=O)c1ccccc1NC(=O)c1ccc(OCCCCCC[n+]2ccc3ccccc3c2)cc1.[Br-]', 'CC(=O)c1ccccc1NC(=O)c1ccc(OCCCCC[n+]2cccc3ccccc32)cc1.[Br-]', 'CC(=O)c1ccccc1NC(=O)c1ccc(OCCCCC[n+]2ccccc2)cc1.[Br-]', 'CC(=O)c1ccccc1NC(=O)c1ccc(OCCCCCC[n+]2ccccc2)cc1.[Br-]', 'CN(CCCCCCCOc1ccc(-c2oc3ccccc3c2C(=O)c2c3ccccc3cc3ccccc23)cc1)Cc1ccccc1', 'CCN(CC)Cc1cccc(C(=O)c2c(-c3ccc(OCCCCCCCN(C)Cc4ccccc4)cc3)oc3ccccc23)c1', 'CCN(CC)Cc1ccc(C(=O)c2c(-c3ccc(OCCCCCCCN(C)Cc4ccccc4)cc3)oc3ccccc23)cc1', 'CCN(CC)CCOc1cccc(C(=O)c2c(-c3ccc(OCCCCCCCN(C)Cc4ccccc4)cc3)oc3ccccc23)c1', 'CCN(CC)CCOc1ccc(C(=O)c2c(-c3ccc(OCCCCCCCN(C)Cc4ccccc4)cc3)oc3ccccc23)cc1', 'CN(CCCCCCCOc1ccc(-c2oc3ccccc3c2C(=O)c2ccc(OCCN3CCOCC3)cc2)cc1)Cc1ccccc1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc(-c3oc4ccccc4c3C(=O)c3ccc(C)cc3)cc2)c1', 'CN(CCCCCCCOc1cccc(-c2oc3ccccc3c2C(=O)c2ccccc2)c1)Cc1ccccc1', 'COc1cc2c(cc1OC)-c1cc3ccc(OC)c(OC)c3c[n+]1CC2.O.[Cl-]', 'COc1cc(/C=C/C(=O)NCCCCCCNc2c3c(nc4ccccc24)CCCC3)ccc1O', 'CO[P@]1(=O)CC[C@@H]2COC(=O)C2=C(C)O1', 'CO[P@]1(=O)CC[C@@H]2COC(=O)C2=C(C)O1', 'CS(=O)(=O)[O-].CS(=O)(=O)[O-].NC(=O)c1cc[n+](COC[n+]2ccc(/C=N/O)cc2/C=N/O)cc1', 'CS(=O)(=O)[O-].CS(=O)(=O)[O-].NC(=O)c1cc[n+](COC[n+]2ccc(/C=N/O)cc2/C=N/O)cc1', 'CS(=O)(=O)[O-].CS(=O)(=O)[O-].NC(=O)c1cc[n+](COC[n+]2ccc(/C=N/O)cc2/C=N/O)cc1', 'CS(=O)(=O)[O-].CS(=O)(=O)[O-].NC(=O)c1cc[n+](COC[n+]2ccc(/C=N/O)cc2/C=N/O)cc1', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCc1ccc(O)c(/C=N/O)n1)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCc1ccc(O)c(/C=N/O)n1)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCc1ccc(O)c(/C=N/O)n1)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCc1ccc(O)c(/C=N/O)n1)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCCc1ccc(O)c(/C=N/O)n1)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCCc1ccc(O)c(/C=N/O)n1)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCCc1ccc(O)c(/C=N/O)n1)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCCc1ccc(O)c(/C=N/O)n1)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCCc1ccc(O)c(/C=N/O)n1)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCCc1ccc(O)c(/C=N/O)n1)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCCCc1ccc(O)c(/C=N/O)n1)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCCCc1ccc(O)c(/C=N/O)n1)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCCCc1ccc(O)c(/C=N/O)n1)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCOCCCc1ccc(O)c(/C=N/O)n1)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCc1ccnc(/C=N/O)c1O)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCCc1ccnc(/C=N/O)c1O)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCCc1ccnc(/C=N/O)c1O)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCCCc1cc(C)nc(/C=N/O)c1O)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCCCc1cc(C)nc(/C=N/O)c1O)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCc1cnc(/C=N/O)c(O)c1)CC2', 'Nc1c2c(nc3c1C(c1ccccc1)c1ccc4ccccc4c1O3)CCCC2', 'Cc1ccc(C2c3ccc4ccccc4c3Oc3nc4c(c(N)c32)CCCC4)cc1', 'Nc1c2c(nc3c1C(c1ccc(O)cc1)c1ccc4ccccc4c1O3)CCCC2', 'COc1ccccc1C1c2ccc3ccccc3c2Oc2nc3c(c(N)c21)CCCC3', 'COc1cc(C2c3ccc4ccccc4c3Oc3nc4c(c(N)c32)CCCC4)ccc1O', 'COc1ccc(C2c3ccc4ccccc4c3Oc3nc4c(c(N)c32)CCCC4)c(OC)c1', 'COc1cc(C2c3ccc4ccccc4c3Oc3nc4c(c(N)c32)CCCC4)cc(OC)c1OC', 'Nc1c2c(nc3c1C(c1ccc([N+](=O)[O-])cc1)c1ccc4ccccc4c1O3)CCCC2', 'Nc1c2c(nc3c1C(c1c(Cl)cccc1Cl)c1ccc4ccccc4c1O3)CCCC2', 'Nc1c2c(nc3c1C(c1ccc(Cl)c(Cl)c1)c1ccc4ccccc4c1O3)CCCC2', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1cccc([N+](=O)[O-])c1)CCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1ccccc1)CCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1cccnc1OC)CCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1cccnc1Br)CCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1cccnc1Cl)CCCC3', 'CC1(C)CC(=O)C2=C(C1)Oc1nc3c(c(N)c1[C@H]2c1ccccc1)CCCC3', 'CC1(C)CC(=O)C2=C(C1)Oc1nc3c(c(N)c1[C@H]2c1ccc(F)cc1)CCCC3', 'CC1(C)CC(=O)C2=C(C1)Oc1nc3c(c(N)c1[C@H]2c1ccccc1C(F)(F)F)CCCC3', 'CC1(C)CC(=O)C2=C(C1)Oc1nc3c(c(N)c1[C@H]2c1cccc([N+](=O)[O-])c1)CCCC3', 'CC1(C)CC(=O)C2=C(C1)Oc1nc3c(c(N)c1[C@H]2c1ccc([N+](=O)[O-])cc1)CCCC3', 'Cc1ccc([C@H]2C3=C(CC(C)(C)CC3=O)Oc3nc4c(c(N)c32)CCCC4)cc1', 'COc1ccccc1[C@H]1C2=C(CC(C)(C)CC2=O)Oc2nc3c(c(N)c21)CCCC3', 'COc1cccc([C@H]2C3=C(CC(C)(C)CC3=O)Oc3nc4c(c(N)c32)CCCC4)c1', 'COc1ccc([C@H]2C3=C(CC(C)(C)CC3=O)Oc3nc4c(c(N)c32)CCCC4)cc1', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2[C@H]1c1ccccc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2[C@H]1c1ccc(F)cc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2[C@H]1c1ccccc1[N+](=O)[O-])CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2[C@H]1c1cccc([N+](=O)[O-])c1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2[C@H]1c1ccc([N+](=O)[O-])cc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2[C@H]1c1ccc(C)cc1)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2[C@@H]1c1ccccc1OC)CCCC3', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2[C@H]1c1cccc(OC)c1)CCCC3', 'CC1(C)CC(=O)C2=C(C1)Nc1nc3c(c(N)c1[C@H]2c1ccccc1)CCCC3', 'CC1(C)CC(=O)C2=C(C1)Nc1nc3c(c(N)c1[C@H]2c1ccc(F)cc1)CCCC3', 'Cc1ccc([C@H]2C3=C(CC(C)(C)CC3=O)Nc3nc4c(c(N)c32)CCCC4)cc1', 'COc1ccc([C@H]2C3=C(CC(C)(C)CC3=O)Nc3nc4c(c(N)c32)CCCC4)cc1', 'CC1(C)CC(=O)C2=C(C1)Nc1nc3c(c(N)c1[C@H]2c1ccncc1)CCCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1ccc(C)cc1)CCCCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1ccccc1OC)CCCCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1ccc(Cl)cc1)CCCCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1cccc([N+](=O)[O-])c1)CCCCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1ccccc1)CCCCCC3', 'CCOC(=O)c1c(C)nc2nc3c(c(N)c2c1-c1ccc(OC)cc1)CCCCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1ccncc1)CCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1cccs1)CCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1ccc(OC)cc1)CCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1ccncc1)CCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1cccnc1)CCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1cccs1)CCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1ccco1)CCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1ccncc1)CCCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1cccnc1)CCCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1cccs1)CCCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1cccc(OC)c1)CCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1ccc(OC)c(OC)c1)CCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1cccc(OC)c1)CCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1ccc(OC)c(OC)c1)CCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1cccc(OC)c1)CCCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1ccccc1OC)CCCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1ccc(OC)c(OC)c1)CCCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1ccccc1)CCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1ccc(C)cc1)CCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1ccc(Cl)cc1)CCCC3', 'CCOC(=O)C1=C(C)Oc2nc3c(c(N)c2[C@H]1c1ccc(C#N)cc1)CCCC3', 'CNCOc1cccc(CN(C)CCCCCCCOc2ccc3c(=O)c4ccccc4oc3c2)c1', 'CCN(CCCCCOc1ccc2c(=O)c3ccccc3oc2c1)Cc1cccc(OC(=O)NC)c1', 'Nc1cc(-c2ccccc2)nc2nc(-c3ccccc3)cc(C(F)(F)F)c12', 'Nc1cc(-c2cccc(F)c2)nc2nc(-c3ccccc3)cc(C(F)(F)F)c12', 'Nc1cc(-c2ccc(F)cc2)nc2nc(-c3ccccc3)cc(C(F)(F)F)c12', 'Nc1cc(-c2ccc(Cl)cc2)nc2nc(-c3ccccc3)cc(C(F)(F)F)c12', 'Nc1cc(-c2ccccc2Br)nc2nc(-c3ccccc3)cc(C(F)(F)F)c12', 'COc1ccc(-c2cc(N)c3c(C(F)(F)F)cc(-c4ccccc4)nc3n2)cc1', 'Nc1cc(-c2cccs2)nc2nc(-c3ccccc3)cc(C(F)(F)F)c12', 'CC(C)Cc1cc(N)c2c(C(F)(F)F)cc(-c3ccccc3)nc2n1', 'Nc1c2c(nc3nc(-c4ccccc4)cc(C(F)(F)F)c13)CCC2', 'Nc1c2c(nc3nc(-c4ccccc4)cc(C(F)(F)F)c13)CCCC2', 'Nc1c2c(nc3nc(-c4ccccc4)cc(C(F)(F)F)c13)CCCCC2', 'Nc1c2c(nc3nc(-c4ccccc4)cc(C(F)(F)F)c13)CCCCCC2', 'CC1Cc2nc3nc(-c4ccccc4)cc(C(F)(F)F)c3c(N)c2C1', 'CC1CCc2c(nc3nc(-c4ccccc4)cc(C(F)(F)F)c3c2N)C1', 'CC1CCc2nc3nc(-c4ccccc4)cc(C(F)(F)F)c3c(N)c2C1', 'Nc1c2c(nc3nc(-c4ccccc4)cc(C(F)(F)F)c13)-c1ccccc1C2', 'Nc1c2c(nc3nc(-c4ccccc4)cc(C(F)(F)F)c13)-c1ccccc1CC2', 'CC1(C)CCCC2(C)OC(=O)C=C12', 'COc1cc(O)c(C2CC(=O)c3c(O)cc(O)cc3O2)cc1C/C=C(\\C)CCC=C(C)C', 'CC(C)=CCC/C(C)=C/Cc1cc(-c2cc(=O)c3c(O)cc(O)cc3o2)c(O)cc1O', 'CC(C)=CCC/C(C)=C/Cc1cc(-c2cc(=O)c3c(O)cc(O)cc3o2)c(O)cc1O', 'COc1cc(O)c(-c2cc(=O)c3c(O)cc(O)cc3o2)cc1C/C=C(\\C)CCC=C(C)C', 'COc1cc(O)c(-c2cc(=O)c3c(O)cc(O)cc3o2)cc1C/C=C(\\C)CCC=C(C)C', 'O=c1c(OC2O[C@H](CO)[C@H](O)[C@H](O)[C@H]2O)c(-c2ccc(O)cc2)oc2cc(O)cc(O)c12', 'COc1cc2occ(-c3ccc(OC4CCN(Cc5ccccc5)CC4)cc3)c(=O)c2cc1OC', 'CC(C)(C)c1ccc(/C(N)=N/OC(=O)c2cnn(-c3ccc([N+](=O)[O-])cc3)c2C(F)(F)F)cc1', 'N/C(=N\\OC(=O)c1cnn(-c2ccc([N+](=O)[O-])cc2)c1C(F)(F)F)c1ccc(Cl)cc1', 'O=C(Nc1ccc(OC(F)(F)F)cc1)N1CCCN(CCc2ccccc2)CC1', 'Cc1c(S(=O)(=O)c2ccc(Cl)cc2)csc1C(=O)NC1CCN(Cc2ccccc2)CC1', 'O=C(CN1CCN(Cc2ccccc2)CC1)Nc1ccc(C(F)(F)F)cc1Cl', 'O=C(Nc1ccc(OC(F)(F)F)cc1)OC1CN(C(c2ccccc2)c2ccccc2)C1', 'CSc1cccc(/N=C2\\NC(=O)C(CC(=O)Nc3ccc(F)c(F)c3)S2)c1', 'O=c1c(Cl)c(N2CCN(c3ccc(C(F)(F)F)cn3)CC2)cnn1-c1c(Cl)cc(C(F)(F)F)cc1Cl', 'O=[N+]([O-])c1ccc(OCC(O)CN2CCCN(c3nc4ccc(Cl)cc4s3)CC2)cc1', 'CC(CN1CCN(Cc2ccccc2)CC1)N(C(=O)c1cccc([N+](=O)[O-])c1)c1ccccn1', 'N/C(=N\\OC(=O)c1cnn(-c2ccc([N+](=O)[O-])cc2)c1C(F)(F)F)c1ccc(C(F)(F)F)cc1', 'Clc1ccc([C@@H]2CC(c3c4ccccc4cc4ccccc34)=NN2)cc1', 'c1ccc([C@@H]2CC(c3c4ccccc4cc4ccccc34)=NN2)cc1', 'Oc1ccccc1[C@@H]1CC(c2c3ccccc3cc3ccccc23)=NN1', 'Oc1ccc([C@@H]2CC(c3c4ccccc4cc4ccccc34)=NN2)c(O)c1', 'c1ccc([C@@H]2CC(c3c4ccccc4cc4ccccc34)=NN2)nc1', 'O=[N+]([O-])c1cccc([C@@H]2CC(c3c4ccccc4cc4ccccc34)=NN2)c1', 'COc1ccc([C@@H]2CC(c3c4ccccc4cc4ccccc34)=NN2)cc1', 'COc1cc2cc(NC(=O)[C@H]3CCC[C@@H](NCc4ccccc4)C3)c(=O)oc2cc1OC', 'COc1cc2cc(NC(=O)[C@H]3CCC[C@@H](NCc4ccccc4)C3)c(=O)oc2cc1OC', 'COc1cc2cc(NC(=O)[C@H]3CCC[C@@H](NCc4ccccc4)C3)c(=O)oc2cc1OC', 'COc1cc2cc(NC(=O)[C@H]3CCC[C@@H](NCc4ccccc4)C3)c(=O)oc2cc1OC', 'COc1cc2cc(NC(=O)[C@H]3CCC[C@@H](NCc4ccccc4)C3)c(=O)oc2cc1OC', 'COc1ccc(/C=N/NC(=O)CN2CCN(Cc3ccccc3)CC2)cc1OC', 'COc1ccc(/C=N/NC(=O)CN2CCN(Cc3ccccc3)CC2)cc1', 'CCOc1ccc(/C=N/NC(=O)CN2CCN(Cc3ccccc3)CC2)cc1', 'COc1ccc(/C=N/NC(=O)CN2CCC(Cc3ccccc3)CC2)cc1OC', 'COc1ccc(/C=N/NC(=O)CN2CCC(Cc3ccccc3)CC2)cc1', 'COc1ccc(/C=N/NC(=O)CN2CCC(Cc3ccccc3)CC2)cc1', 'CCOc1ccc(/C=N/NC(=O)CN2CCC(Cc3ccccc3)CC2)cc1', 'CCOc1ccc(/C=N/NC(=O)CN2CCC(Cc3ccccc3)CC2)cc1', 'CCOc1ccc(/C=N/NC(=O)CN2CCC(Cc3ccccc3)CC2)cc1', 'Cc1ccc(/C=C2\\CN(C(=O)C[C@H]3C[C@H]4CCCN4[C@]34C(=O)Nc3ccccc34)C[C@@]3(C[C@H]4CCCN4[C@@]34C(=O)Nc3ccccc34)C2=O)cc1', 'O=C(C[C@H]1C[C@H]2CCCN2[C@]12C(=O)Nc1ccccc12)N1C/C(=C\\c2ccc(Cl)cc2Cl)C(=O)[C@]2(C[C@H]3CCCN3[C@@]23C(=O)Nc2ccccc23)C1', 'O=C(C[C@H]1C[C@H]2CCCN2[C@]12C(=O)Nc1ccccc12)N1C/C(=C\\c2cccc([N+](=O)[O-])c2)C(=O)[C@]2(C[C@H]3CCCN3[C@@]23C(=O)Nc2ccccc23)C1', 'O=C(C[C@H]1C[C@H]2CCCN2[C@]12C(=O)Nc1ccccc12)N1C/C(=C\\c2ccccc2F)C(=O)[C@]2(C[C@H]3CCCN3[C@@]23C(=O)Nc2ccccc23)C1', 'O=C(C[C@H]1C[C@H]2CCCN2[C@]12C(=O)Nc1ccccc12)N1C/C(=C\\c2ccccc2Cl)C(=O)[C@]2(C[C@H]3CCCN3[C@@]23C(=O)Nc2ccccc23)C1', 'C=CC(=O)N1C/C(=C\\c2ccccc2Cl)C(=O)/C(=C/c2ccccc2Cl)C1', 'C=CC(=O)N1C/C(=C\\c2ccccc2Cl)C(=O)/C(=C/c2ccccc2Cl)C1', 'C=CC(=O)N1C/C(=C\\c2ccccc2F)C(=O)/C(=C/c2ccccc2F)C1', 'C=CC(=O)N1C/C(=C\\c2ccccc2F)C(=O)/C(=C/c2ccccc2F)C1', 'C=CC(=O)N1C/C(=C\\c2cccc([N+](=O)[O-])c2)C(=O)/C(=C/c2cccc([N+](=O)[O-])c2)C1', 'C=CC(=O)N1C/C(=C\\c2cccc([N+](=O)[O-])c2)C(=O)/C(=C/c2cccc([N+](=O)[O-])c2)C1', 'C=CC(=O)N1C/C(=C\\c2ccccc2C)C(=O)/C(=C/c2ccccc2C)C1', 'C=CC(=O)N1C/C(=C\\c2ccccc2C)C(=O)/C(=C/c2ccccc2C)C1', 'C=CC(=O)N1C/C(=C\\c2ccccc2OC)C(=O)/C(=C/c2ccccc2OC)C1', 'C=CC(=O)N1C/C(=C\\c2ccccc2OC)C(=O)/C(=C/c2ccccc2OC)C1', 'O=C(C[C@H]1C[C@H]2CCCN2[C@]12C(=O)Nc1ccccc12)N1C/C(=C\\c2ccc(F)cc2)C(=O)[C@]2(C[C@H]3CCCN3[C@@]23C(=O)Nc2ccccc23)C1', 'O=C(C[C@H]1C[C@H]2CCCN2[C@]12C(=O)Nc1ccccc12)N1C/C(=C\\c2ccc(Cl)cc2)C(=O)[C@]2(C[C@H]3CCCN3[C@@]23C(=O)Nc2ccccc23)C1', 'COc1ccccc1/C=C1\\CN(C(=O)C[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2ccccc23)C[C@@]2(C[C@H]3CCCN3[C@@]23C(=O)Nc2ccccc23)C1=O', 'Cc1ccccc1/C=C1\\CN(C(=O)C[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2ccccc23)C[C@@]2(C[C@H]3CCCN3[C@@]23C(=O)Nc2ccccc23)C1=O', 'O=C(C[C@H]1C[C@H]2CCCN2[C@]12C(=O)Nc1ccccc12)N1C/C(=C\\c2ccccc2)C(=O)[C@]2(C[C@H]3CCCN3[C@@]23C(=O)Nc2ccccc23)C1', 'O=C1/C(=C/c2cccc3ccccc23)CN(C(=O)C[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2ccccc23)C/C1=C\\c1cccc2ccccc12', 'C=CC(=O)N1C/C(=C\\c2ccc(Cl)cc2Cl)C(=O)/C(=C/c2ccc(Cl)cc2Cl)C1', 'C=CC(=O)N1C/C(=C\\c2ccc(Cl)cc2Cl)C(=O)/C(=C/c2ccc(Cl)cc2Cl)C1', 'C=CC(=O)N1C/C(=C\\c2cccc3ccccc23)C(=O)/C(=C/c2cccc3ccccc23)C1', 'C=CC(=O)N1C/C(=C\\c2cccc3ccccc23)C(=O)/C(=C/c2cccc3ccccc23)C1', 'O=C1/C(=C/c2ccccc2)CN(C(=O)C[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2ccccc23)C/C1=C\\c1ccccc1', 'Cc1ccccc1/C=C1\\CN(C(=O)C[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2ccccc23)C/C(=C\\c2ccccc2C)C1=O', 'COc1ccccc1/C=C1\\CN(C(=O)C[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2ccccc23)C/C(=C\\c2ccccc2OC)C1=O', 'O=C1/C(=C/c2ccccc2Cl)CN(C(=O)C[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2ccccc23)C/C1=C\\c1ccccc1Cl', 'O=C1/C(=C/c2ccccc2F)CN(C(=O)C[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2ccccc23)C/C1=C\\c1ccccc1F', 'O=C1/C(=C/c2cccc([N+](=O)[O-])c2)CN(C(=O)C[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2ccccc23)C/C1=C\\c1cccc([N+](=O)[O-])c1', 'O=C1/C(=C/c2ccc(Cl)cc2Cl)CN(C(=O)C[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2ccccc23)C/C1=C\\c1ccc(Cl)cc1Cl', 'Cc1ccc(/C=C2\\CN(C(=O)C[C@H]3C[C@H]4CCCN4[C@]34C(=O)Nc3ccccc34)C/C(=C\\c3ccc(C)cc3)C2=O)cc1', 'O=C1/C(=C/c2ccc(Cl)cc2)CN(C(=O)C[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2ccccc23)C/C1=C\\c1ccc(Cl)cc1', 'O=C1/C(=C/c2ccc(F)cc2)CN(C(=O)C[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2ccccc23)C/C1=C\\c1ccc(F)cc1', 'O=C(C[C@H]1C[C@H]2CCCN2[C@]12C(=O)Nc1ccccc12)N1C/C(=C\\c2cccc3ccccc23)C(=O)[C@]2(C[C@H]3CCCN3[C@@]23C(=O)Nc2ccccc23)C1', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3)CC1)C2.Cl.O', 'C/C=C(\\C)C(=O)N[C@H]1CC[C@@]2(C)C(=CC[C@H]3[C@@H]4CC[C@H]([C@H](C)N(C)C)[C@@]4(C)CC[C@@H]32)[C@H]1O', 'Fc1cccc(Cc2nnc3sc(-c4ccoc4)nn23)c1', 'COc1ccc(-c2nnc3sc(-c4ccoc4)nn23)cc1', 'Brc1ccc(-c2nnc3sc(-c4ccoc4)nn23)cc1', 'Brc1cccc(-c2nnc3sc(-c4ccoc4)nn23)c1', 'Clc1ccccc1-c1nnc2sc(-c3ccoc3)nn12', 'Clc1ccc(-c2nnc3sc(-c4ccoc4)nn23)cc1', 'Fc1ccccc1Cc1nnc2sc(-c3ccoc3)nn12', 'Fc1ccc(Cc2nnc3sc(-c4ccoc4)nn23)cc1', 'COc1ccccc1-c1nnc2sc(-c3ccoc3)nn12', 'COc1cccc(-c2nnc3sc(-c4ccoc4)nn23)c1', 'O=C1/C(=C/c2cccc3ccccc23)CN(C(=O)[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2cc(Cl)ccc23)C/C1=C\\c1cccc2ccccc12', 'O=C1/C(=C/c2ccccc2)CN(C(=O)[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2cc(Cl)ccc23)C/C1=C\\c1ccccc1', 'Cc1ccccc1/C=C1\\CN(C(=O)[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2cc(Cl)ccc23)C/C(=C\\c2ccccc2C)C1=O', 'COc1ccccc1/C=C1\\CN(C(=O)[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2cc(Cl)ccc23)C/C(=C\\c2ccccc2OC)C1=O', 'O=C1/C(=C/c2ccccc2Cl)CN(C(=O)[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2cc(Cl)ccc23)C/C1=C\\c1ccccc1Cl', 'O=C1/C(=C/c2ccccc2F)CN(C(=O)[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2cc(Cl)ccc23)C/C1=C\\c1ccccc1F', 'O=C1/C(=C/c2cccc([N+](=O)[O-])c2)CN(C(=O)[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2cc(Cl)ccc23)C/C1=C\\c1cccc([N+](=O)[O-])c1', 'O=C1/C(=C/c2ccc(Cl)cc2Cl)CN(C(=O)[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2cc(Cl)ccc23)C/C1=C\\c1ccc(Cl)cc1Cl', 'Cc1ccc(/C=C2\\CN(C(=O)[C@H]3C[C@H]4CCCN4[C@]34C(=O)Nc3cc(Cl)ccc34)C/C(=C\\c3ccc(C)cc3)C2=O)cc1', 'O=C1/C(=C/c2ccc(Cl)cc2)CN(C(=O)[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2cc(Cl)ccc23)C/C1=C\\c1ccc(Cl)cc1', 'O=C1/C(=C/c2ccc(F)cc2)CN(C(=O)[C@H]2C[C@H]3CCCN3[C@]23C(=O)Nc2cc(Cl)ccc23)C/C1=C\\c1ccc(F)cc1', 'COc1ccc2cc(C(=O)CN3CCN(Cc4ccccc4)CC3)c(=O)oc2c1', 'CN(C)C(=O)[C@H]1C[C@@H]1c1ccc(-c2ncn(C)c2Sc2ccc(Cl)cn2)cc1', 'Cc1cc(=O)oc2cc(OCCN3CCN(CC(=O)Nc4c5c(nc6ccccc46)CCCC5)CC3)ccc12', 'CN(C)CCCN(c1nc2ccccc2[nH]1)c1nc2ccccc2[nH]1', 'c1ccc2[nH]c(N3CCCCC3)nc2c1', 'CC1CCCCN1c1nc2ccccc2[nH]1', 'CC1=C[C@H]2Cc3[nH]c(=O)ccc3[C@@]3(C1)[C@@H]2[C@H](Cl)CCN3C', 'CC1=C[C@H]2Cc3[nH]c(=O)ccc3[C@@]3(C1)[C@@H]2[C@H](C)CN3C', 'C=C[C@@H]1[C@H]2C=C(C)C[C@]1(N)c1ccc(=O)[nH]c1C2', 'CC(C)N(CCOc1ccc(-c2nc3ccccc3[nH]2)cc1)C(C)C', 'c1ccc2[nH]c(-c3ccc(OCCN4CCCC4)cc3)nc2c1', 'c1ccc2[nH]c(-c3ccc(OCCN4CCCCC4)cc3)nc2c1', 'Cc1ccc2[nH]c(-c3ccc(OCCN(C)C)cc3)nc2c1', 'CCN(CC)CCOc1ccc(-c2nc3cc(C)ccc3[nH]2)cc1', 'Cc1ccc2[nH]c(-c3ccc(OCCN4CCCC4)cc3)nc2c1', 'Cc1ccc2[nH]c(-c3ccc(OCCN4CCCCC4)cc3)nc2c1', 'Cc1ccc2[nH]c(-c3ccc(OCCN4CCOCC4)cc3)nc2c1', 'CN(C)CCOc1ccc(-c2nc3cc(Cl)ccc3[nH]2)cc1', 'CCN(CC)CCOc1ccc(-c2nc3cc(Cl)ccc3[nH]2)cc1', 'CC(C)N(CCOc1ccc(-c2nc3cc(Cl)ccc3[nH]2)cc1)C(C)C', 'Clc1ccc2[nH]c(-c3ccc(OCCN4CCCC4)cc3)nc2c1', 'Clc1ccc2[nH]c(-c3ccc(OCCN4CCCCC4)cc3)nc2c1', 'Clc1ccc2[nH]c(-c3ccc(OCCN4CCOCC4)cc3)nc2c1', 'CN(C)CCOc1ccc(-c2nc3ccccc3[nH]2)cc1', 'N#Cc1c(Cl)nc(NC2CCN(Cc3ccccc3)CC2)c(C#N)c1-c1ccccc1', 'N#Cc1c(Cl)nc(NCCC2CCN(Cc3ccccc3)CC2)c(C#N)c1-c1ccccc1', 'N#Cc1c(Cl)nc(NCCCC2CCN(Cc3ccccc3)CC2)c(C#N)c1-c1ccccc1', 'N#Cc1c(Cl)nc(NCCCCC2CCN(Cc3ccccc3)CC2)c(C#N)c1-c1ccccc1', 'N#Cc1cc(C#N)c(NC2CCN(Cc3ccccc3)CC2)nc1Cl', 'N#Cc1cc(C#N)c(NCCC2CCN(Cc3ccccc3)CC2)nc1Cl', 'N#Cc1cc(C#N)c(NCCCC2CCN(Cc3ccccc3)CC2)nc1Cl', 'N#Cc1cc(C#N)c(NCCCCC2CCN(Cc3ccccc3)CC2)nc1Cl', 'N#Cc1c(NCCC2CCN(Cc3ccccc3)CC2)nc(NCCC2CCN(Cc3ccccc3)CC2)c(C#N)c1-c1ccccc1', 'N#Cc1cc(C#N)c(NC2CCN(Cc3ccccc3)CC2)nc1NC1CCN(Cc2ccccc2)CC1', 'O=C(NNC=C1C(=O)NC(=S)NC1=O)Nc1ccccc1', 'O=C1NC(=S)NC(=O)C1=CNNC(=S)Nc1ccc(F)cc1', 'O=C1NC(=S)NC(=O)C1=CNNC(=S)Nc1ccc(Br)cc1', 'O=C1NC(=O)C(=CNNC(=S)Nc2cccc(C(F)(F)F)c2)C(=O)N1', 'O=C1NC(=S)NC(=O)C1=CNNC(=S)Nc1ccc(C(F)(F)F)cc1', 'CC(=O)NC=C1C(=O)NC(=O)NC1=O', 'COc1cccc(CCc2cccc(OC)[n+]2C)[n+]1C.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1cccc(CCc2cccc(OC)[n+]2C)[n+]1C.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1cccc(CCCCCc2cccc(OC)[n+]2C)[n+]1C.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1cccc(CCCCCc2cccc(OC)[n+]2C)[n+]1C.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1cccc(CCCCCCc2cccc(OC)[n+]2C)[n+]1C.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1cccc(CCCCCCc2cccc(OC)[n+]2C)[n+]1C.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1cccc(CCCCCCCc2cccc(OC)[n+]2C)[n+]1C.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1cccc(CCCCCCCc2cccc(OC)[n+]2C)[n+]1C.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1cccc(CCCCCCCCc2cccc(OC)[n+]2C)[n+]1C.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1cccc(CCCCCCCCc2cccc(OC)[n+]2C)[n+]1C.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1cc2c(cc1OC)C[N+](C)(CCCCc1cccc(OC)[n+]1C)CC2.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1cc2c(cc1OC)C[N+](C)(CCCCc1cccc(OC)[n+]1C)CC2.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1cc2c(cc1OC)C[N+](C)(CCCCCc1cccc(OC)[n+]1C)CC2.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1cc2c(cc1OC)C[N+](C)(CCCCCc1cccc(OC)[n+]1C)CC2.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1cc2c(cc1OC)C[N+](C)(CCCCCCc1cccc(OC)[n+]1C)CC2.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1cc2c(cc1OC)C[N+](C)(CCCCCCc1cccc(OC)[n+]1C)CC2.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1cc2c(cc1OC)C[N+](C)(CCCCCCCc1cccc(OC)[n+]1C)CC2.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1cc2c(cc1OC)C[N+](C)(CCCCCCCc1cccc(OC)[n+]1C)CC2.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1cc2c(cc1OC)C(=O)C(CCCCCc1cccc(OC)[n+]1C)C2.O=S(=O)([O-])C(F)(F)F', 'COc1cc2c(cc1OC)C(=O)C(CCCCCc1cccc(OC)[n+]1C)C2.O=S(=O)([O-])C(F)(F)F', 'COc1cc2c(cc1OC)C(=O)C(CCCCCCc1cccc(OC)[n+]1C)C2.O=S(=O)([O-])C(F)(F)F', 'COc1cc2c(cc1OC)C(=O)C(CCCCCCc1cccc(OC)[n+]1C)C2.O=S(=O)([O-])C(F)(F)F', 'COc1cc2c(cc1OC)C(=O)C(CCCCCCCc1cccc(OC)[n+]1C)C2.O=S(=O)([O-])C(F)(F)F', 'COc1cc2c(cc1OC)C(=O)C(CCCCCCCc1cccc(OC)[n+]1C)C2.O=S(=O)([O-])C(F)(F)F', 'COc1ccc(C(F)(F)F)c[n+]1CCCCCO.O=S(=O)([O-])C(F)(F)F', 'COc1ccc(C(F)(F)F)c[n+]1CCC[n+]1cc(C(F)(F)F)ccc1OC.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1ccc(C(F)(F)F)c[n+]1CCC[n+]1cc(C(F)(F)F)ccc1OC.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1ccc(C(F)(F)F)c[n+]1CCCCC[n+]1cc(C(F)(F)F)ccc1OC.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1ccc(C(F)(F)F)c[n+]1CCCCC[n+]1cc(C(F)(F)F)ccc1OC.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1ccc(C(F)(F)F)c[n+]1CCCOc1ccc2c(C)cc(=O)oc2c1.O=S(=O)([O-])C(F)(F)F', 'COc1ccc(C(F)(F)F)c[n+]1CCCOc1ccc2c(C)cc(=O)oc2c1.O=S(=O)([O-])C(F)(F)F', 'COc1ccc(C(F)(F)F)c[n+]1CCCCOc1ccc2c(C)cc(=O)oc2c1.O=S(=O)([O-])C(F)(F)F', 'COc1ccc(C(F)(F)F)c[n+]1CCCCOc1ccc2c(C)cc(=O)oc2c1.O=S(=O)([O-])C(F)(F)F', 'COc1ccc(C(F)(F)F)c[n+]1CCCCCOc1ccc2c(C)cc(=O)oc2c1.O=S(=O)([O-])C(F)(F)F', 'COc1ccc(C(F)(F)F)c[n+]1CCCCCOc1ccc2c(C)cc(=O)oc2c1.O=S(=O)([O-])C(F)(F)F', 'COc1c(F)c[n+](CCCOc2ccc3c(C)cc(=O)oc3c2)cc1F.O=S(=O)([O-])C(F)(F)F', 'COc1c(F)c[n+](CCCOc2ccc3c(C)cc(=O)oc3c2)cc1F.O=S(=O)([O-])C(F)(F)F', 'COc1c(F)c[n+](CCCCOc2ccc3c(C)cc(=O)oc3c2)cc1F.O=S(=O)([O-])C(F)(F)F', 'COc1c(F)c[n+](CCCCOc2ccc3c(C)cc(=O)oc3c2)cc1F.O=S(=O)([O-])C(F)(F)F', 'COc1c(F)c[n+](CCCCCOc2ccc3c(C)cc(=O)oc3c2)cc1F.O=S(=O)([O-])C(F)(F)F', 'COc1c(F)c[n+](CCCCCOc2ccc3c(C)cc(=O)oc3c2)cc1F.O=S(=O)([O-])C(F)(F)F', 'COc1cccc[n+]1C.F[B-](F)(F)F', 'COc1cccc[n+]1C.F[B-](F)(F)F', 'COc1c(F)ccc[n+]1C.F[B-](F)(F)F', 'COc1ccc(F)c[n+]1C.F[B-](F)(F)F', 'COc1ccc(F)c[n+]1C.F[B-](F)(F)F', 'COc1ccc(C(F)(F)F)c[n+]1C.F[B-](F)(F)F', 'COc1ccc(C(F)(F)F)c[n+]1C.F[B-](F)(F)F', 'COc1cccc(C(F)(F)F)[n+]1C.F[B-](F)(F)F', 'COc1cccc(C(F)(F)F)[n+]1C.F[B-](F)(F)F', 'COc1ccc([N+](=O)[O-])c[n+]1C.F[B-](F)(F)F', 'COc1ccc([N+](=O)[O-])c[n+]1C.F[B-](F)(F)F', 'COc1cc(C#N)cc[n+]1C.F[B-](F)(F)F', 'COc1cc(C#N)cc[n+]1C.F[B-](F)(F)F', 'COc1ccc(C#N)c[n+]1C.F[B-](F)(F)F', 'COc1ccc(C#N)c[n+]1C.F[B-](F)(F)F', 'COc1cccc(C#N)[n+]1C.F[B-](F)(F)F', 'COc1cccc(C#N)[n+]1C.F[B-](F)(F)F', 'COc1c(F)c[n+](C)cc1F.O=S(=O)([O-])C(F)(F)F', 'COc1c(F)c[n+](C)cc1F.O=S(=O)([O-])C(F)(F)F', 'COc1c(F)c[n+](CCC[n+]2cc(F)c(OC)c(F)c2)cc1F.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'COc1c(F)c[n+](CCC[n+]2cc(F)c(OC)c(F)c2)cc1F.O=S(=O)([O-])C(F)(F)F.O=S(=O)([O-])C(F)(F)F', 'Cl.Cl.O=C(CCc1ccc2nc(-c3ccccc3)c3c(c2c1)OCCC3)NCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'CCCCCCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN(CCC)C1C2.Cc1ccc(C(=O)OC(C(=O)O)C(OC(=O)c2ccc(C)cc2)C(=O)O)cc1.O', 'C=CC(C)(C)c1cc([C@@H]2CC(=O)c3c(O)cc(O)c(CC=C(C)C)c3O2)c(O)cc1O', 'CCn1c(=O)n(C)s/c1=N\\CCC1CCN(Cc2ccccc2)CC1', 'COc1cc2[se]n(CCC3CCN(Cc4ccccc4)CC3)c(=O)c2cc1OC', 'COc1cc2[se]n(CCCC3CCN(Cc4ccccc4)CC3)c(=O)c2cc1OC', 'COc1cc2[se]n(CCCCC3CCN(Cc4ccccc4)CC3)c(=O)c2cc1OC', 'COc1cc2c(cc1OC)[Se](=O)N(CCC1CCN(Cc3ccccc3)CC1)C2=O', 'CC(C)Cc1ccc(C(C)C(=O)NCCCNCCCNC2=CC(=O)C(NCCCNCCCNC(=O)C(C)c3ccc(CC(C)C)cc3)=CC2=O)cc1', 'CC1=C(CC(=O)NCCCNCCCNC2=CC(=O)C(NCCCNCCCNC(=O)CC3=C(C)/C(=C\\c4ccc([S+](C)[O-])cc4)c4ccc(F)cc43)=CC2=O)c2cc(F)ccc2/C1=C/c1ccc([S+](C)[O-])cc1', 'O=C(Cc1ccccc1Nc1c(Cl)cccc1Cl)NCCCNCCCNC1=CC(=O)C(NCCCNCCCNC(=O)Cc2ccccc2Nc2c(Cl)cccc2Cl)=CC1=O', 'O=C(Cc1ccccc1Nc1c(Cl)cccc1Cl)NCCCNCCCNC1=CC(=O)C(NCCCNCCCNC(=O)Cc2ccccc2Nc2c(Cl)cccc2Cl)=CC1=O', 'COc1ccc2c(c1)c(CC(=O)NCCCNCCCNC1=CC(=O)C(NCCCNCCCNC(=O)Cc3c(C)n(C(=O)c4ccc(Cl)cc4)c4ccc(OC)cc34)=CC1=O)c(C)n2C(=O)c1ccc(Cl)cc1', 'C[C@@H](C(=O)NCCCNCCCNC1=CC(=O)C(NCCCNCCCNC(=O)[C@H](C)c2ccc(-c3ccccc3)c(F)c2)=CC1=O)c1ccc(-c2ccccc2)c(F)c1', 'C[C@H](C(=O)NCCCNCCCNC1=CC(=O)C(NCCCNCCCNC(=O)[C@@H](C)c2ccc(-c3ccccc3)c(F)c2)=CC1=O)c1ccc(-c2ccccc2)c(F)c1', 'COc1ccc(/C=C2\\COc3cc(OCCCCCNc4c5c(nc6ccccc46)CCCC5)ccc3C2=O)cc1', 'COc1ccc(/C=C2\\COc3cc(OCCCCCCNc4c5c(nc6ccccc46)CCCC5)ccc3C2=O)cc1', 'COc1ccc(/C=C2\\COc3cc(OCCCCCCCCNc4c5c(nc6ccccc46)CCCC5)ccc3C2=O)cc1', 'O=C1/C(=C/c2ccc(F)cc2)COc2cc(OCCCCCCNc3c4c(nc5ccccc35)CCCC4)ccc21', 'O=c1c2ccccc2[se]n1CCCCCCNc1c2c(nc3ccccc13)CCCC2', 'COc1cc2[se]n(CCCCCCNc3c4c(nc5ccccc35)CCCC4)c(=O)c2cc1OC', '[Br-].[Br-].c1cc[n+](CCCCCCC[n+]2ccc3ccccc3c2)cc1', '[Br-].[Br-].c1cc[n+](CCCCCCCC[n+]2ccc3ccccc3c2)cc1', '[Br-].[Br-].c1cc[n+](CCCCCCCCC[n+]2ccc3ccccc3c2)cc1', '[Br-].[Br-].c1cc[n+](CCCCCCCCCC[n+]2ccc3ccccc3c2)cc1', '[Br-].[Br-].c1cc[n+](CCCCCCCCCCC[n+]2ccc3ccccc3c2)cc1', '[Br-].[Br-].c1cc[n+](CCCCCCC[n+]2cccc3ccccc32)cc1', '[Br-].[Br-].c1cc[n+](CCCCCCCC[n+]2cccc3ccccc32)cc1', '[Br-].[Br-].c1cc[n+](CCCCCCCCC[n+]2cccc3ccccc32)cc1', '[Br-].[Br-].c1cc[n+](CCCCCCCCCC[n+]2cccc3ccccc32)cc1', '[Br-].[Br-].c1cc[n+](CCCCCCCCCC[n+]2cccc3ccccc32)cc1', '[Br-].[Br-].c1cc[n+](CCCCCCCCCC[n+]2cccc3ccccc32)cc1', '[Br-].[Br-].c1cc[n+](CCCCCCCCCCC[n+]2cccc3ccccc32)cc1', '[Br-].[Br-].c1cc[n+](CCCCCCCCCCCC[n+]2cccc3ccccc32)cc1', '[Br-].[Br-].c1cc[n+](CCCCCCCCCCCC[n+]2cccc3ccccc32)cc1', '[Br-].[Br-].c1cc[n+](CCCCCCCCCCCC[n+]2cccc3ccccc32)cc1', 'O=[N+]([O-])c1cccc2c(NCCCCCCCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc12)CCCC3', 'Nc1cccc2c(NCCCCCCCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc12)CCCC3', 'O=C(CC[C@H](NC(=O)CNC(=O)OCc1ccccc1)C(=O)N[C@@H](Cc1ccccc1)C(=O)OCc1ccccc1)NCC(=O)Nc1cccc2c(NCCCCCCCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc12)CCCC3', 'CN(CCCCNc1c2c(nc3ccccc13)CCCC2)CCCNc1c2c(nc3c(NC(=O)CNC(=O)CC[C@H](NC(=O)CNC(=O)OCc4ccccc4)C(=O)N[C@@H](Cc4ccccc4)C(=O)OCc4ccccc4)cccc13)CCCC2', 'CN(CCCCNc1c2c(nc3c(NC(=O)CNC(=O)CC[C@H](NC(=O)CNC(=O)OCc4ccccc4)C(=O)N[C@@H](Cc4ccccc4)C(=O)OCc4ccccc4)cccc13)CCCC2)CCCNc1c2c(nc3ccccc13)CCCC2', 'COc1cc2[se]n(-c3ccc(CN4CCCC4)cc3)c(=O)c2cc1OC', 'COc1cc2[se]n(-c3ccc(CN4CCCCC4)cc3)c(=O)c2cc1OC', 'COc1cc2[se]n(-c3ccc(CCN4CCCC4)cc3)c(=O)c2cc1OC', 'COc1cc2[se]n(-c3ccc(CCN4CCCCC4)cc3)c(=O)c2cc1OC', 'COc1cc2[se]n(-c3ccc(OCCN4CCCCC4)cc3)c(=O)c2cc1OC', 'COc1cc2[se]n(-c3ccc(OCCCN4CCCCC4)cc3)c(=O)c2cc1OC', 'COc1cc2[se]n(-c3ccc(OCCCCN4CCCCC4)cc3)c(=O)c2cc1OC', 'O=c1c2ccccc2[se]n1-c1ccc(CN2CCCCC2)cc1', 'O=c1c2ccccc2[se]n1-c1ccc(OCCN2CCCCC2)cc1', 'COc1ccccc1/C=C1\\CN(C(=O)[C@@H]2CCN(C)[C@]23C(=O)Nc2ccccc23)C[C@]2(C1=O)[C@H](c1ccccc1OC)CN(C)[C@]21C(=O)Nc2ccccc21', 'CN1CC[C@@H](C(=O)N2C/C(=C\\c3ccccc3Cl)C(=O)[C@@]3(C2)[C@H](c2ccccc2Cl)CN(C)[C@]32C(=O)Nc3ccccc32)[C@@]12C(=O)Nc1ccccc12', 'CN1CC[C@@H](C(=O)N2C/C(=C\\c3ccccc3F)C(=O)[C@@]3(C2)[C@H](c2ccccc2F)CN(C)[C@]32C(=O)Nc3ccccc32)[C@@]12C(=O)Nc1ccccc12', 'CN1CC[C@@H](C(=O)N2C/C(=C\\c3cccc([N+](=O)[O-])c3)C(=O)[C@@]3(C2)[C@H](c2cccc([N+](=O)[O-])c2)CN(C)[C@]32C(=O)Nc3ccccc32)[C@@]12C(=O)Nc1ccccc12', 'CN1CC[C@@H](C(=O)N2C/C(=C\\c3ccc(Cl)cc3Cl)C(=O)[C@@]3(C2)[C@H](c2ccc(Cl)cc2Cl)CN(C)[C@]32C(=O)Nc3ccccc32)[C@@]12C(=O)Nc1ccccc12', 'Cc1ccc(/C=C2\\CN(C(=O)[C@@H]3CCN(C)[C@]34C(=O)Nc3ccccc34)C[C@]3(C2=O)[C@H](c2ccc(C)cc2)CN(C)[C@]32C(=O)Nc3ccccc32)cc1', 'CN1CC[C@@H](C(=O)N2C/C(=C\\c3ccc(Cl)cc3)C(=O)[C@@]3(C2)[C@H](c2ccc(Cl)cc2)CN(C)[C@]32C(=O)Nc3ccccc32)[C@@]12C(=O)Nc1ccccc12', 'CN1CC[C@@H](C(=O)N2C/C(=C\\c3ccc(F)cc3)C(=O)[C@@]3(C2)[C@H](c2ccc(F)cc2)CN(C)[C@]32C(=O)Nc3ccccc32)[C@@]12C(=O)Nc1ccccc12', 'CN1CC[C@@H](C(=O)N2C/C(=C\\c3ccc(F)cc3)C(=O)[C@@]3(C2)[C@H](c2ccc(F)cc2)CN(C)[C@]32C(=O)Nc3ccccc32)[C@@]12C(=O)Nc1ccccc12', 'CN1CC[C@@H](C(=O)N2C/C(=C\\c3cccc4ccccc34)C(=O)[C@@]3(C2)[C@H](c2cccc4ccccc24)CN(C)[C@]32C(=O)Nc3ccccc32)[C@@]12C(=O)Nc1ccccc12', 'C=CC(=O)N1C/C(=C\\c2ccccc2)C(=O)[C@@]2(C1)[C@H](c1ccccc1)CN(C)[C@]21C(=O)Nc2ccccc21', 'C=CC(=O)N1C/C(=C\\c2ccccc2C)C(=O)[C@@]2(C1)[C@H](c1ccccc1C)CN(C)[C@]21C(=O)Nc2ccccc21', 'C=CC(=O)N1C/C(=C\\c2ccccc2OC)C(=O)[C@@]2(C1)[C@H](c1ccccc1OC)CN(C)[C@]21C(=O)Nc2ccccc21', 'C=CC(=O)N1C/C(=C\\c2ccccc2Cl)C(=O)[C@@]2(C1)[C@H](c1ccccc1Cl)CN(C)[C@]21C(=O)Nc2ccccc21', 'C=CC(=O)N1C/C(=C\\c2ccccc2F)C(=O)[C@@]2(C1)[C@H](c1ccccc1F)CN(C)[C@]21C(=O)Nc2ccccc21', 'C=CC(=O)N1C/C(=C\\c2cccc([N+](=O)[O-])c2)C(=O)[C@@]2(C1)[C@H](c1cccc([N+](=O)[O-])c1)CN(C)[C@]21C(=O)Nc2ccccc21', 'C=CC(=O)N1C/C(=C\\c2ccc(Cl)cc2Cl)C(=O)[C@@]2(C1)[C@H](c1ccc(Cl)cc1Cl)CN(C)[C@]21C(=O)Nc2ccccc21', 'C=CC(=O)N1C/C(=C\\c2ccc(C)cc2)C(=O)[C@@]2(C1)[C@H](c1ccc(C)cc1)CN(C)[C@]21C(=O)Nc2ccccc21', 'C=CC(=O)N1C/C(=C\\c2ccc(Cl)cc2)C(=O)[C@@]2(C1)[C@H](c1ccc(Cl)cc1)CN(C)[C@]21C(=O)Nc2ccccc21', 'C=CC(=O)N1C/C(=C\\c2ccc(F)cc2)C(=O)[C@@]2(C1)[C@H](c1ccc(F)cc1)CN(C)[C@]21C(=O)Nc2ccccc21', 'C=CC(=O)N1C/C(=C\\c2cccc3ccccc23)C(=O)[C@@]2(C1)[C@H](c1cccc3ccccc13)CN(C)[C@]21C(=O)Nc2ccccc21', 'CN1CC[C@@H](C(=O)N2C/C(=C\\c3ccccc3)C(=O)[C@@]3(C2)[C@H](c2ccccc2)CN(C)[C@]32C(=O)Nc3ccccc32)[C@@]12C(=O)Nc1ccccc12', 'Cc1ccccc1/C=C1\\CN(C(=O)[C@@H]2CCN(C)[C@]23C(=O)Nc2ccccc23)C[C@]2(C1=O)[C@H](c1ccccc1C)CN(C)[C@]21C(=O)Nc2ccccc21', 'c1ccc2c(c1)c(OCC1CCCCC1)nn2CC1CCCCC1', 'c1ccc(COc2nn(Cc3ccccc3)c3ccccc23)cc1', 'CCCCCOc1nn(CCCCC)c2ccc([N+](=O)[O-])cc12', 'CCCCCn1nc(OCc2ccc(OC)cc2)c2cc([N+](=O)[O-])ccc21', 'COc1ccc(COc2nn(CCN(C(C)C)C(C)C)c3ccc([N+](=O)[O-])cc23)cc1', 'O=[N+]([O-])c1ccc2c(c1)c(OCc1cccc3ccccc13)nn2Cc1cccc2ccccc12', 'CCCCCn1nc(OCc2ccc3ccccc3c2)c2cc([N+](=O)[O-])ccc21', 'O=[N+]([O-])c1ccc2c(c1)c(OCc1ccc3ccccc3c1)nn2Cc1ccccc1', 'Nc1ccc2c(c1)c(OCc1ccccc1)nn2Cc1ccccc1', 'CCOC(=O)c1ccc2nc(-c3cccnc3)c3c(c2c1)N(Cc1ccccc1)CCC3', 'CCNC(=O)c1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)N(Cc1ccccc1)CCC3', 'CCNC(=O)c1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)N(Cc1ccccc1)CCC3', 'CCNC(=O)c1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)N(Cc1ccccc1)CCC3', 'CCNCc1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)N(Cc1ccccc1)CCC3', 'CCOC(=O)c1ccc2nc(-c3cccnc3)c3c(c2c1)NCCC3', 'CCNC(=O)c1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)NCCC3', 'CCNC(=O)c1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)NCCC3', 'CCNC(=O)c1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)NCCC3', 'CCNC(=O)c1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)NCCC3', 'CCNCc1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)NCCC3', 'CCNCc1ccc2nc(-c3cccnc3)c3c(c2c1)NCCC3', 'CCOC(=O)c1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)N(Cc1ccccc1)C(=O)CC3', 'CCOC(=O)c1ccc2nc(-c3cccnc3)c3c(c2c1)N(Cc1ccccc1)C(=O)CC3', 'CCN(CCCCOc1ccc(-c2coc3cc(O)cc(O)c3c2=O)cc1)Cc1ccccc1OC', 'CCN(CCCCOc1ccc(-c2coc3cc(OCCCCN(CC)Cc4ccccc4)cc(O)c3c2=O)cc1)Cc1ccccc1', 'CCN(CCCCOc1ccc(-c2coc3cc(OCCCCN(CC)Cc4ccccc4OC)cc(O)c3c2=O)cc1)Cc1ccccc1OC', 'CCN(CCCCCCOc1cc(O)c2c(=O)c(-c3ccc(O)cc3)coc2c1)Cc1ccccc1', 'Cn1ncc2c1-c1sccc1C2NC(=O)CN1CCN(Cc2ccccc2)CC1', 'Cn1ncc2c1-c1sccc1C2NC(=O)CN1CCN(Cc2ccccc2F)CC1', 'O=C(CN1CCN(Cc2ccccc2)CC1)NC1c2cn[nH]c2-c2c(Br)sc(Br)c21', 'Cn1ncc2c1-c1c(Br)sc(Br)c1C2NC(=O)CN1CCN(Cc2ccccc2)CC1', 'O=C(CN1CCN(Cc2ccccc2)CC1)NC1c2cnn(-c3ccccc3)c2-c2c(Br)sc(Br)c21', 'Cn1ncc2c1-c1c(Br)sc(Br)c1C2NC(=O)CN1CCN(Cc2ccccc2F)CC1', 'COc1cc2c(cc1OC)C(NC(=O)CN1CCN(Cc3ccccc3)CC1)c1cnn(C)c1-2', 'COc1cc2c(cc1OC)C(NC(=O)CN1CCN(Cc3ccccc3)CC1)c1cnn(-c3ccccc3)c1-2', 'COc1cc2c(cc1OC)C(NC(=O)CN1CCN(Cc3ccccc3F)CC1)c1cnn(C)c1-2', 'COc1cc2c(cc1OC)C(NC(=O)CN1CCN(Cc3ccccc3F)CC1)c1cnn(-c3ccccc3)c1-2', 'O=C(CN1CCN(Cc2ccccc2)CC1)NC1c2cn[nH]c2-c2sccc21', 'C[Si](C)(C)c1cccc(C(=O)C(F)(F)F)c1', 'C[Si](C)(C)c1cccc(C(=O)C(F)(F)F)c1', 'CNC(=C[N+](=O)[O-])NCCSCc1csc(CN(C)C)n1', 'Cl.Cl.c1ccc(CN2CCC(CCNc3ccc(-c4ccccc4)nn3)CC2)cc1', 'Cl.Cl.c1ccc(CN2CCC(CCNc3ccc(-c4ccccc4)nn3)CC2)cc1', 'Cl.Cl.Cl.Cl.O=C(NCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)c1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)OCCC3', 'Cl.Cl.Cl.O=C(CCc1ccc2nc(-c3ccccc3)c3c(c2c1)OCCC3)NCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Cl.Cl.Cl.O=C(CCc1ccc2nc(-c3ccccc3)c3c(c2c1)OCCC3)NCCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Cl.Cl.Oc1c(CNCCCCCCCCCNc2c3c(nc4ccccc24)CCCC3)ccc2cccnc12', 'Cl.Cl.Cl.Cl.O=C(NCCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)c1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)OCCC3', 'Cl.Cl.Cl.Cl.O=C(NCCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)c1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)OCCC3', 'Cl.Cl.Cl.Cl.O=C(NCCCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)c1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)OCCC3', 'Cl.Cl.Cl.Cl.O=C(NCCCCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)c1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)OCCC3', 'Cc1ccc2ccc(CNCCCCCCCCNc3c4c(nc5ccccc35)CCCC4)c(O)c2n1.Cl.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCNc4c5c(nc6ccccc46)CCCC5)c3C(C1)C2.Cl.Cl.Cl', 'Cl.Cl.Cl.O=C(CCc1ccc2nc(-c3ccccc3)c3c(c2c1)OCCC3)NCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCNc4c5c(nc6ccccc46)CCCC5)c3C(C1)C2.Cl.Cl.Cl', 'COc1cc2c(cc1OC)CC(CC1CCN(CCNc3c4c(nc5ccccc35)CCCC4)CC1)C2.Cl.Cl', 'COc1cc2c(cc1OC)CC(CC1CCN(CCCNc3c4c(nc5ccccc35)CCCC4)CC1)C2.Cl.Cl', 'COc1cc2c(cc1OC)CC(CC1CCN(CCNc3c4c(nc5cc(Cl)ccc35)CCCC4)CC1)C2.Cl.Cl.Cl', 'Cl.Cl.Cl.O=C(CCc1ccc2nc(-c3ccccc3)c3c(c2c1)OCCC3)NCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(CCNc3c4c(nc5cc(Cl)ccc35)CCCC4)CC1)C2.Cl.Cl.Cl', 'Cl.Cl.Cl.Cl.O=C(NCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)c1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)OCCC3', 'Cl.Cl.Cl.Oc1c(CNCCCCCCCCCNc2c3c(nc4ccccc24)CCCC3)cc(Cl)c2cccnc12', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCCNc4c5c(nc6ccccc46)CCCC5)c3C(C1)C2.Cl.Cl.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)c3C(C1)C2.Cl.Cl.Cl.Cl', 'CC1=C[C@@H]2Cc3nc4cc(Cl)ccc4c(NCCCCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)c3[C@H](C1)C2.Cl.Cl.Cl.Cl', 'CC1=C[C@H]2Cc3nc4cc(Cl)ccc4c(NCCCCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)c3[C@@H](C1)C2.Cl.Cl.Cl.Cl', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(CCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)CC1)C2.Cl.Cl.Cl', 'CC1=C[C@@H]2Cc3nc4cc(Cl)ccc4c(NCCCCCCCNc4c5c(nc6ccccc46)CCCC5)c3[C@H](C1)C2.Cl.Cl.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)c3C(C1)C2.Cl.Cl.Cl.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)c3C(C1)C2.Cl.Cl.Cl.Cl', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(CCCNc3c4c(nc5ccccc35)CCCC4)CC1)C2.Cl.Cl', 'COc1cc2c(cc1OC)CC(CC1CCN(CCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)CC1)C2.Cl.Cl.Cl', 'CC1=C[C@H]2Cc3nc4cc(Cl)ccc4c(NCCCCCCCNc4c5c(nc6ccccc46)CCCC5)c3[C@@H](C1)C2.Cl.Cl.Cl', 'Cc1ccc(C(=O)Nc2cc3cc(C)ccc3oc2=O)cc1', 'COc1ccc(C(=O)Nc2cc3ccccc3oc2=O)cc1', 'O=C(Nc1cc2ccccc2oc1=O)c1ccc(Cl)cc1', 'O=C(Nc1cc2ccccc2oc1=O)c1ccc([N+](=O)[O-])cc1', 'O=C(Nc1cc2ccccc2oc1=O)c1ccc(Cl)c(Cl)c1', 'C[C@@]12CC[C@@H]3[C@@](CC[C@H]4[C@@]3(C)CCC[C@@]4(C)C(=O)OCC[N+](C)(C)Cc3ccccc3)(CC1=O)C2.[Br-]', 'C[C@@]12CC[C@@H]3[C@@](CC[C@H]4[C@@]3(C)CCC[C@@]4(C)C(=O)OCC[N+](C)(C)CCCCCCCCCC[N+](C)(C)CCOC(=O)[C@]3(C)CCC[C@@]4(C)[C@@H]5CC[C@@]6(C)C[C@]5(CC[C@@H]43)CC6=O)(CC1=O)C2.[Br-].[Br-]', 'C[C@@]12CC[C@@H]3[C@@](CC[C@H]4[C@@]3(C)CCC[C@@]4(C)C(=O)OCC[N+](C)(C)CCCCCCCCCCCC[N+](C)(C)CCOC(=O)[C@]3(C)CCC[C@@]4(C)[C@@H]5CC[C@@]6(C)C[C@]5(CC[C@@H]43)CC6=O)(CC1=O)C2.[Br-].[Br-]', 'C[C@@]12CC[C@@H]3[C@]4(CC[C@H]5[C@@]3(C)CCC[C@@]5(C)C(=O)OCC[N+](C)(C)CCCCCCCCC[N+](C)(C)CCOC(=O)[C@]3(C)CCC[C@@]5(C)[C@@H]6CC[C@@]7(C)C[C@]6(CC[C@@H]53)C[C@H]7OC(=O)CCCCCCCCC(=O)O[C@@H]1C4)C2.[Br-].[Br-]', 'CC[N+](CC)(CC)CCOCCOCCOC(=O)[C@]1(C)CCC[C@@]2(C)[C@@H]3CC[C@@]4(C)C[C@]3(CC[C@@H]21)CC4=O.[Br-]', 'C[C@H]1C[C@@]23CC[C@H]4[C@@](C)(CCC[C@@]4(C)C(=O)OCC[N+](C)(C)C)[C@@H]2CC[C@]1(Cl)C3.[I-]', 'C[C@H]1C[C@@]23CC[C@H]4[C@@](C)(CCC[C@@]4(C)C(=O)OCC[N+](C)(C)Cc4ccccc4)[C@@H]2CC[C@]1(Cl)C3.[Br-]', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCNC(=O)c4cc(O)c5c(c4)C(=O)c4cccc(O)c4C5=O)c3C(C1)C2.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCNC(=O)c4cc(O)c5c(c4)C(=O)c4cccc(O)c4C5=O)c3C(C1)C2.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCNC(=O)c4cc(O)c5c(c4)C(=O)c4cccc(O)c4C5=O)c3C(C1)C2.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCNC(=O)c4cc(O)c5c(c4)C(=O)c4cccc(O)c4C5=O)c3C(C1)C2.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCNC(=O)c4cc(O)c5c(c4)C(=O)c4cccc(O)c4C5=O)c3C(C1)C2.Cl', 'CC1=C[C@H]2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCNC(=O)c4cc(O)c5c(c4)C(=O)c4cccc(O)c4C5=O)c3[C@@H](C1)C2.Cl', 'CC1=C[C@@H]2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCNC(=O)c4cc(O)c5c(c4)C(=O)c4cccc(O)c4C5=O)c3[C@H](C1)C2.Cl', 'CC1=C[C@@H]2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCNC(=O)c4cc(O)c5c(c4)C(=O)c4cccc(O)c4C5=O)c3[C@H](C1)C2.Cl', 'CC1=C[C@@H]2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCNC(=O)c4cc(O)c5c(c4)C(=O)c4cccc(O)c4C5=O)c3[C@H](C1)C2.Cl', 'COc1cc2[nH]c(=O)c(C(=O)NCC3CCN(Cc4ccccc4)CC3)c(O)c2cc1OC', 'O=C(NCC1CCN(Cc2ccccc2)CC1)c1c(O)c2cc(O)c(O)cc2[nH]c1=O', 'COc1ccc(C(C#N)N2CCCC2)cc1OC', 'COc1ccc(C(C#N)N2CCOCC2)cc1OC', 'N#CC(c1ccc2c(c1)OCO2)N1CCCCC1', 'N#CC(c1ccc2c(c1)OCO2)N1CCCC1', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCCNC(=O)c4cc(O)c5c(c4)C(=O)c4cccc(O)c4C5=O)c3C(C1)C2.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCCCNC(=O)c4cc(O)c5c(c4)C(=O)c4cccc(O)c4C5=O)c3C(C1)C2.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCc4ccc(CNC(=O)c5cc(O)c6c(c5)C(=O)c5cccc(O)c5C6=O)cc4)c3C(C1)C2.Cl', 'NC(=O)c1cc[n+](CCC[n+]2ccccc2/C=N/O)cc1.[Cl-].[Cl-]', 'CN(CCNc1c2c(nc3ccccc13)CCCC2)Cc1ccnc(/C(N)=N/O)c1O', 'CN(CCCNc1c2c(nc3ccccc13)CCCC2)Cc1ccnc(/C(N)=N/O)c1O', 'O/N=C/c1nc(CCCCCNc2c3c(nc4ccccc24)CCCC3)ccc1O', 'O/N=C/c1nc(CCCCCNc2c3c(nc4ccccc24)CCCC3)ccc1O', 'O/N=C/c1nc(CCCCCNc2c3c(nc4ccccc24)CCCC3)ccc1O', 'O/N=C/c1nc(CCCCCNc2c3c(nc4ccccc24)CCCC3)ccc1O', 'O/N=C/c1nc(CCCCNc2c3c(nc4ccccc24)CCCC3)ccc1O', 'O/N=C/c1nc(CCCCNc2c3c(nc4ccccc24)CCCC3)ccc1O', 'O/N=C/c1nc(CCCCNc2c3c(nc4ccccc24)CCCC3)ccc1O', 'O/N=C/c1nc(CCCCNc2c3c(nc4ccccc24)CCCC3)ccc1O', 'O/N=C/c1nc(CCCCNc2c3c(nc4ccccc24)CCCC3)ccc1O', 'O/N=C/c1nc(CCCCNc2c3c(nc4ccccc24)CCCC3)ccc1O', 'O/N=C/c1ncc(CCCCCNc2c3c(nc4ccccc24)CCCC3)cc1O', 'N/C(=N\\O)c1nc(CCCCCNc2c3c(nc4ccccc24)CCCC3)ccc1O', 'O/N=C/c1nccc(CCCCCNc2c3c(nc4ccccc24)CCCC3)c1O', 'c1ccc(CCCCCNc2c3c(nc4ccccc24)CCCC3)nc1', 'CC(=O)/C(=N\\O)C(=O)NCC1CCN(Cc2ccccc2)CC1', 'Cl.O=C(/C=N/O)NCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(/C=N/O)NCCCCNc1c2c(nc3ccccc13)CCCC2', 'N#C/C(=N\\O)C(=O)NCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(/C=N/O)NCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(/C=N/O)c1cccnc1OCC1CCN(Cc2ccccc2)CC1', 'O=C(/C=N/O)c1ccc(OCC2CCN(Cc3ccccc3)CC2)nc1', 'O=C(/C=N/O)c1cnc(OCC2CCN(Cc3ccccc3)CC2)nc1', 'O=C(CCCOc1cccc2[nH]c3ccccc3c12)NCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(CCCOc1cccc2[nH]c3ccccc3c12)NCCCNc1c2c(nc3ccccc13)CCCC2', 'c1ccc2c(NCCSSCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCSSCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'c1ccc2c(NCCSSCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'CN(Cc1ccccc1)Cc1ccc(C(=O)c2ccc(O)c(F)c2)cc1', 'O=C(c1ccc(CN2CCC(CCO)CC2)cc1)c1ccc(O)c(F)c1', 'CC[C@H]1[C@H](C[C@H](C)C(=O)OC)O[C@]23CC[C@@H]([C@@H]4C[C@H](C)C(=O)O4)N2CCCC[C@H]13', 'CC[C@H]1[C@H]2OC(=O)[C@](C)(O)[C@H]2[C@]23CC[C@@H]([C@@H]4C[C@@H](C)C(=O)O4)N2CCCC[C@@H]13', 'CC[C@@H]1[C@H]2OC(=O)[C@@](C)(O)[C@H]2[C@]23CC[C@@H]([C@@H]4C[C@H](C)C(=O)O4)N2CCCC[C@H]13', 'CC[C@H]1[C@@H]2OC(=O)[C@@](C)(O)[C@@H]2[C@@]23CC[C@@H]([C@@H]4C[C@H](C)C(=O)O4)N2CCCC[C@H]13', 'O=C(/C=N/O)Nc1ccc[n+](CC[n+]2cccc(NC(=O)/C=N/O)c2)c1.[Br-].[Br-]', 'O=C(/C=N/O)Nc1ccc[n+](CC[n+]2cccc(NC(=O)/C=N/O)c2)c1.[Br-].[Br-]', 'O=C(/C=N\\O)Nc1ccc[n+](CCC[n+]2cccc(NC(=O)/C=N\\O)c2)c1.[Br-].[Br-]', 'O=C(/C=N\\O)Nc1ccc[n+](CCC[n+]2cccc(NC(=O)/C=N\\O)c2)c1.[Br-].[Br-]', 'O=C(/C=N\\O)Nc1ccc[n+](CCCC[n+]2cccc(NC(=O)/C=N\\O)c2)c1.[Br-].[Br-]', 'O=C(/C=N\\O)Nc1ccc[n+](CCCC[n+]2cccc(NC(=O)/C=N\\O)c2)c1.[Br-].[Br-]', 'O=C(/C=N\\O)Nc1ccc[n+](CCCCC[n+]2cccc(NC(=O)/C=N\\O)c2)c1.[Br-].[Br-]', 'O=C(/C=N\\O)Nc1ccc[n+](CCCCC[n+]2cccc(NC(=O)/C=N\\O)c2)c1.[Br-].[Br-]', 'O=C(/C=N\\O)Nc1ccc[n+](CCCCCC[n+]2cccc(NC(=O)/C=N\\O)c2)c1.[Br-].[Br-]', 'O=C(/C=N\\O)Nc1ccc[n+](CCCCCC[n+]2cccc(NC(=O)/C=N\\O)c2)c1.[Br-].[Br-]', 'O=C(/C=N\\O)Nc1ccc[n+](CCCCCCC[n+]2cccc(NC(=O)/C=N\\O)c2)c1.[Br-].[Br-]', 'O=C(/C=N\\O)Nc1ccc[n+](CCCCCCC[n+]2cccc(NC(=O)/C=N\\O)c2)c1.[Br-].[Br-]', 'O=C(/C=N/O)Nc1ccc[n+](Cc2ccccc2C[n+]2cccc(NC(=O)/C=N/O)c2)c1.[Br-].[Br-]', 'O=C(/C=N/O)Nc1ccc[n+](Cc2ccccc2C[n+]2cccc(NC(=O)/C=N/O)c2)c1.[Br-].[Br-]', 'O=C(/C=N/O)Nc1ccc[n+](Cc2cccc(C[n+]3cccc(NC(=O)/C=N/O)c3)c2)c1.[Br-].[Br-]', 'O=C(/C=N/O)Nc1ccc[n+](Cc2cccc(C[n+]3cccc(NC(=O)/C=N/O)c3)c2)c1.[Br-].[Br-]', 'O=C(/C=N/O)Nc1ccc[n+](Cc2ccc(C[n+]3cccc(NC(=O)/C=N/O)c3)cc2)c1.[Br-].[Br-]', 'O=C(/C=N/O)Nc1ccc[n+](Cc2ccc(C[n+]3cccc(NC(=O)/C=N/O)c3)cc2)c1.[Br-].[Br-]', 'CC(C)Nc1nc(NCc2ccccc2)c2ccccc2n1', 'CCN(CC)CCCCCCOc1ccc(CC2NCCc3cc(OC)c(OC)cc32)cc1', 'COc1cc2c(cc1OC)C(Cc1ccc(OCCCCCCN(C)C)cc1)NCC2', 'CCCN(CCC)CCCCCCOc1ccc(CC2NCCc3cc(OC)c(OC)cc32)cc1', 'CCCCN(CCCC)CCCCCCOc1ccc(CC2NCCc3cc(OC)c(OC)cc32)cc1', 'COc1cc2c(cc1OC)C(Cc1ccc(OCCCCCCN3CCCC3C)cc1)NCC2', 'COc1cc2c(cc1OC)C(Cc1ccc(OCCCCCCN3CCCCC3C)cc1)NCC2', 'COc1cc2c(cc1OC)C(Cc1ccc(OCCCCCCN3CCCCC3)cc1)NCC2', 'COc1cc2c(cc1OC)C(Cc1ccc(OCCCCCCN3CCCC3)cc1)NCC2', 'COc1cc2c(cc1OC)C(Cc1ccc(OCCCCCCN3CCC(O)CC3)cc1)NCC2', 'COc1cc2c(cc1OC)C(Cc1ccc(OCCCCCCN3CCOCC3)cc1)NCC2', 'COc1cc2c(cc1OC)C(Cc1ccc(OCCCCCCN3CCC(O)(Cc4ccccc4)CC3)cc1)NCC2', 'COc1cc2c(cc1OC)C(Cc1ccc(OCCCCCCN3CCN(c4ccccc4)CC3)cc1)NCC2', 'COc1cc2c(cc1OC)C(Cc1ccc(OCCCCCCN3CCN(Cc4ccccc4)CC3)cc1)NCC2', 'COc1cc2c(cc1OC)C(Cc1ccc(OCCCCCCN3CCC(c4ccccc4)CC3)cc1)NCC2', 'CN(Cc1ccccc1)Cc1ccc(C(=O)NCCc2c[nH]c3ccccc23)cc1', 'CN(Cc1ccccc1)Cc1ccc(C(=O)NCCc2c[nH]c3ccc(O)cc23)cc1', 'COc1ccc2[nH]cc(CCNC(=O)c3ccc(CN(C)Cc4ccccc4)cc3)c2c1', 'COc1ccc2c(CCNC(=O)c3ccc(CN(C)Cc4ccccc4)cc3)c[nH]c2c1', 'CN(Cc1ccccc1)Cc1ccc(C(=O)NCCc2c[nH]c3cc(F)ccc23)cc1', 'CN(Cc1ccc(C(=O)NCCc2c[nH]c3ccccc23)cc1)Cc1ccccc1Cl', 'COc1ccc2[nH]cc(CCNC(=O)c3ccc(CN(C)Cc4ccccc4Cl)cc3)c2c1', 'CN(Cc1ccc(C(=O)NCCc2c[nH]c3ccccc23)cc1)Cc1cccc(Cl)c1', 'COc1ccc2[nH]cc(CCNC(=O)c3ccc(CN(C)Cc4cccc(Cl)c4)cc3)c2c1', 'COc1ccc2c(CCNC(=O)c3ccc(CN(C)Cc4cccc(Cl)c4)cc3)c[nH]c2c1', 'COc1ccccc1CN(C)Cc1ccc(C(=O)NCCc2c[nH]c3ccccc23)cc1', 'COc1ccc2[nH]cc(CCNC(=O)c3ccc(CN(C)Cc4ccccc4OC)cc3)c2c1', 'COc1cccc(CN(C)Cc2ccc(C(=O)NCCc3c[nH]c4ccccc34)cc2)c1', 'COc1cccc(CN(C)Cc2ccc(C(=O)NCCc3c[nH]c4ccc(OC)cc34)cc2)c1', 'CCOC(OCC)c1ccc(/C=C2\\CCC/C(=C\\c3ccc(C(OCC)OCC)cc3)C2=O)cc1', 'CCOC(OCC)c1ccc(/C=C2\\COC/C(=C\\c3ccc(C(OCC)OCC)cc3)C2=O)cc1', 'CCOC(OCC)c1ccc(/C=C2\\CNC/C(=C\\c3ccc(C(OCC)OCC)cc3)C2=O)cc1', 'CCOC(OCC)c1ccc(/C=C2\\CN(C)C/C(=C\\c3ccc(C(OCC)OCC)cc3)C2=O)cc1', 'CCOC(OCC)c1ccc(/C=C2\\CN(Cc3ccccc3)C/C(=C\\c3ccc(C(OCC)OCC)cc3)C2=O)cc1', 'CCOC(OCC)c1ccc(/C=C/C(=O)/C=C/c2ccc(C(OCC)OCC)cc2)cc1', 'CCOC(OCC)c1ccc(/C=C2\\CC/C(=C\\c3ccc(C(OCC)OCC)cc3)C2=O)cc1', 'CCOC(OCC)c1ccc(/C=C/C(=O)C2CCc3ccccc3C2=O)cc1', 'CN(C)c1ccc(/C=C2/CCC/C(=C\\c3ccc(N(C)C)cc3[N+](=O)[O-])C2=O)c([N+](=O)[O-])c1', 'CN(C)c1ccc(/C=C2/COC/C(=C\\c3ccc(N(C)C)cc3[N+](=O)[O-])C2=O)c([N+](=O)[O-])c1', 'CN(C)c1ccc(/C=C2/CNC/C(=C\\c3ccc(N(C)C)cc3[N+](=O)[O-])C2=O)c([N+](=O)[O-])c1', 'CN1C/C(=C/c2ccc(N(C)C)cc2[N+](=O)[O-])C(=O)/C(=C/c2ccc(N(C)C)cc2[N+](=O)[O-])C1', 'CN(C)c1ccc(/C=C2\\CCc3ccccc3C2=O)c([N+](=O)[O-])c1', 'CN(C)c1ccc(/C=C/C(=O)C2CCc3ccccc3C2=O)cc1', 'O=C1/C(=C\\c2ccc(N3CCCC3)cc2)CNC/C1=C\\c1ccc(N2CCCC2)cc1', 'CN1C/C(=C/c2ccc(N3CCCC3)cc2)C(=O)/C(=C/c2ccc(N3CCCC3)cc2)C1', 'O=C1/C(=C/c2ccc(N3CCCC3)cc2)CC/C1=C\\c1ccc(N2CCCC2)cc1', 'C#CCN(Cc1ccc(O)c2ncccc12)C(C#N)CCC1CCN(Cc2ccccc2)CC1', 'C#CCN(C)Cc1ccc(O)c2ncccc12', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCCNC(=O)c4cc(O)c5c(c4)C(=O)c4cccc(O)c4C5=O)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCc4ccc(CNC(=O)c5cc(O)c6c(c5)C(=O)c5cccc(O)c5C6=O)cc4)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCCCNC(=O)c4cc(O)c5c(c4)C(=O)c4cccc(O)c4C5=O)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCNC(=O)c4cc(O)c5c(c4)C(=O)c4cccc(O)c4C5=O)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCNC(=O)c4cc(O)c5c(c4)C(=O)c4cccc(O)c4C5=O)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCNC(=O)c4cc(O)c5c(c4)C(=O)c4cccc(O)c4C5=O)c3C(C1)C2', 'CC1(C)C[C@@H](c2ccccc2)C(CN2C(=O)c3ccccc3C2=O)=NO1', 'COC(=O)C(CC1=NO[C@@H]2CCC[C@@H]2[C@H]1c1ccccc1)C(=O)OC', 'CC(=O)N1OC(C)(C)CC(c2ccccc2)=C1CN1C(=O)c2ccccc2C1=O', 'COC(=O)C(CC1=C(c2ccc(OC)cc2)CC(C)(C)ON1C(C)=O)C(=O)OC', 'COC(=O)C(CC1=C(c2ccccc2)CC(C)(C)ON1C(C)=O)C(=O)OC', 'COC(=O)C(CC1=C(c2ccc(Cl)cc2)CC(C)(C)ON1C(C)=O)C(=O)OC', 'O=C(C[n+]1ccccc1)N/N=C/c1ccc(/C=N/NC(=O)C[n+]2ccccc2)cc1.[Cl-].[Cl-]', 'COc1cc2oc(-c3ccc(CN4CCCC4)cc3)cc(=O)c2cc1OC', 'O=c1cc(-c2ccc(CN3CCCCC3)cc2)oc2ccccc12', 'Cn1ncc2cc(C(=O)Nc3ccc(Cl)c(Cl)c3)ccc21', 'CCN(CCC(=O)NCCNc1c2c(nc3ccccc13)CCCC2)C1CCCCC1', 'CCN(CCC(=O)NCCCNc1c2c(nc3ccccc13)CCCC2)C1CCCCC1', 'CCN(CCC(=O)NCCCNc1c2c(nc3ccccc13)CCCC2)C1CCCCC1', 'CCN(CCC(=O)NCCCNc1c2c(nc3ccccc13)CCCC2)C1CCCCC1', 'CCN(CCC(=O)NCCCCNc1c2c(nc3ccccc13)CCCC2)C1CCCCC1', 'CCN(CCC(=O)NCCCCNc1c2c(nc3ccccc13)CCCC2)C1CCCCC1', 'CCN(CCC(=O)NCCCCNc1c2c(nc3ccccc13)CCCC2)C1CCCCC1', 'CN(CCC(=O)NCCNc1c2c(nc3ccccc13)CCCC2)C1CCCCC1', 'CN(CCC(=O)NCCCNc1c2c(nc3ccccc13)CCCC2)C1CCCCC1', 'CN(CCC(=O)NCCCCNc1c2c(nc3ccccc13)CCCC2)C1CCCCC1', 'O=C(Cc1cc(=O)oc2cc(O)ccc12)NCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(Cc1cc(=O)oc2cc(O)ccc12)NCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(Cc1cc(=O)oc2cc(O)ccc12)NCCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(Cc1cc(=O)oc2cc(O)ccc12)NCCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(Cc1cc(=O)oc2cc(O)ccc12)NCCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(Cc1cc(=O)oc2cc(O)ccc12)NNC(=S)Nc1c2c(nc3ccccc13)CCCC2', 'O=C(Cc1cc(=O)oc2cc(O)ccc12)NNC(=S)Nc1c2c(nc3ccccc13)CCCC2', 'O=C(Cc1cc(=O)oc2cc(O)ccc12)NNC(=S)Nc1c2c(nc3ccccc13)CCCC2', 'O=C(Cc1cc(=O)oc2cc(O)ccc12)N/N=C1\\SCC(=O)N1c1c2c(nc3ccccc13)CCCC2', 'O=C(Cc1cc(=O)oc2cc(O)ccc12)N/N=C1\\SCC(=O)N1c1c2c(nc3ccccc13)CCCC2', 'O=C(Cc1cc(=O)oc2cc(O)ccc12)N/N=C1\\SCC(=O)N1c1c2c(nc3ccccc13)CCCC2', 'Cl.O=C(NCCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)c1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)NCCC3', 'Cl.O=C(NCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)c1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)NCCC3', 'Cl.O=C(NCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)c1ccc2nc(-c3ccc(Cl)cc3)c3c(c2c1)NCCC3', 'c1cc2c(cc1OCCCN1CCCCC1)CN1CCCC1=N2', 'c1cc2c(cc1OCCCN1CCCCC1)CN1CCCC1=N2', 'c1cc2c(cc1OCCCN1CCCCC1)CN1CCCC1=N2', 'c1cc2c(cc1OCCCN1CCCCC1)CN1CCCC1=N2', 'O=c1c2cc(OCCCN3CCCCC3)ccc2nc2n1Cc1ccccc1-2', 'O=c1c2cc(OCCCN3CCCCC3)ccc2nc2n1Cc1ccccc1-2', 'CCN(CC)CCCOc1ccc2nc3n(c(=O)c2c1)Cc1ccccc1-3', 'CCN(CC)CCCOc1ccc2nc3n(c(=O)c2c1)Cc1ccccc1-3', 'O=c1c2cc(OCCCN3CCCCC3)ccc2nc2n1CCc1ccccc1-2', 'O=c1c2cc(OCCCN3CCCCC3)ccc2nc2n1CCc1ccccc1-2', 'CCN(CC)CCCOc1ccc2nc3n(c(=O)c2c1)CCc1ccccc1-3', 'CCN(CC)CCCOc1ccc2nc3n(c(=O)c2c1)CCc1ccccc1-3', 'O=c1c2cc(OCCCN3CCCCC3)ccc2nc2n1CCCc1ccccc1-2', 'O=c1c2cc(OCCCN3CCCCC3)ccc2nc2n1CCCc1ccccc1-2', 'CCN(CC)CCCOc1ccc2nc3n(c(=O)c2c1)CCCc1ccccc1-3', 'CCN(CC)CCCOc1ccc2nc3n(c(=O)c2c1)CCCc1ccccc1-3', 'O=c1c2cc(OCCCn3ccnc3)ccc2nc2n1CCCc1ccccc1-2', 'O=c1c2cc(OCCCn3ccnc3)ccc2nc2n1CCCc1ccccc1-2', 'O=c1c2cc(OCCCN3CCCCC3)ccc2nc2n1CCC2', 'O=c1c2cc(OCCCN3CCCCC3)ccc2nc2n1CCC2', 'O=c1c2cc(OCCCN3CCCCC3)ccc2nc2n1CCC2', 'O=c1c2cc(OCCCN3CCCCC3)ccc2nc2n1CCC2', 'CCN(CC)CCCOc1ccc2nc3n(c(=O)c2c1)CCC3', 'CCN(CC)CCCOc1ccc2nc3n(c(=O)c2c1)CCC3', 'O=c1c2cc(OCCCN3CCCCC3)ccc2nc2n1CCCC2', 'O=c1c2cc(OCCCN3CCCCC3)ccc2nc2n1CCCC2', 'CCN(CC)CCCOc1ccc2nc3n(c(=O)c2c1)CCCC3', 'CCN(CC)CCCOc1ccc2nc3n(c(=O)c2c1)CCCC3', 'O=c1c2cc(OCCCN3CCCCC3)ccc2nc2n1CCCCC2', 'O=c1c2cc(OCCCN3CCCCC3)ccc2nc2n1CCCCC2', 'CCN(CC)CCCOc1ccc2nc3n(c(=O)c2c1)CCCCC3', 'CCN(CC)CCCOc1ccc2nc3n(c(=O)c2c1)CCCCC3', 'O=C(NCCCCCNc1c2c(nc3ccccc13)CCCC2)c1cc2ccccc2oc1=O', 'COc1ccc2cc(C(=O)NCCCCCNc3c4c(nc5ccccc35)CCCC4)c(=O)oc2c1', 'COc1ccc2oc(=O)c(C(=O)NCCCCCNc3c4c(nc5ccccc35)CCCC4)cc2c1', 'Cc1ccc2oc(=O)c(C(=O)NCCCCCNc3c4c(nc5ccccc35)CCCC4)cc2c1', 'COc1cc2cc(C(=O)NCCCCCNc3c4c(nc5ccccc35)CCCC4)c(=O)oc2cc1OC', 'O=C(NCCCCCNc1c2c(nc3ccccc13)CCCC2)c1cc2cc(OC(F)(F)F)ccc2oc1=O', 'O=C(NCCCCCCNc1c2c(nc3ccccc13)CCCC2)c1cc2ccccc2oc1=O', 'COc1ccc2cc(C(=O)NCCCCCCNc3c4c(nc5ccccc35)CCCC4)c(=O)oc2c1', 'COc1ccc2oc(=O)c(C(=O)NCCCCCCNc3c4c(nc5ccccc35)CCCC4)cc2c1', 'Cc1ccc2oc(=O)c(C(=O)NCCCCCCNc3c4c(nc5ccccc35)CCCC4)cc2c1', 'COc1cc2cc(C(=O)NCCCCCCNc3c4c(nc5ccccc35)CCCC4)c(=O)oc2cc1OC', 'O=C(NCCCCCCNc1c2c(nc3ccccc13)CCCC2)c1cc2cc(OC(F)(F)F)ccc2oc1=O', 'O=C(NCCCCCCCNc1c2c(nc3ccccc13)CCCC2)c1cc2ccccc2oc1=O', 'COc1ccc2cc(C(=O)NCCCCCCCNc3c4c(nc5ccccc35)CCCC4)c(=O)oc2c1', 'COc1ccc2oc(=O)c(C(=O)NCCCCCCCNc3c4c(nc5ccccc35)CCCC4)cc2c1', 'Cc1ccc2oc(=O)c(C(=O)NCCCCCCCNc3c4c(nc5ccccc35)CCCC4)cc2c1', 'COc1cc2cc(C(=O)NCCCCCCCNc3c4c(nc5ccccc35)CCCC4)c(=O)oc2cc1OC', 'O=C(NCCCCCCCNc1c2c(nc3ccccc13)CCCC2)c1cc2cc(OC(F)(F)F)ccc2oc1=O', 'CC(c1ccc2c(c1)oc(=O)c1ccccc12)N(C)C.Cl', 'C[C@@H](c1ccc2c(c1)oc(=O)c1ccccc12)N(C)C.Cl', 'COc1cc(C(C)N(C)C)cc2oc(=O)c3ccccc3c12.Cl', 'COc1cc(C(C)N(C)C)c2c(c1)oc(=O)c1ccccc12.Cl', 'CC(c1cc(O)c2c(c1)oc(=O)c1ccccc12)N(C)C.Cl', 'CC(c1cc(O)cc2oc(=O)c3ccccc3c12)N(C)C.Cl', 'CCN(C)C(=O)Oc1cc(C(C)N(C)C)cc2oc(=O)c3ccccc3c12.Cl', 'CN(CCOc1ccc2c3c(c(=O)oc2c1)CCCC3)Cc1ccccc1', 'COc1cccc(CN(C)CCOc2ccc3c4c(c(=O)oc3c2)CCCC4)c1', 'O=c1oc2cc(OCCN3CCC(Cc4ccccc4)CC3)ccc2c2c1CCCC2', 'COc1cccc(CC2CCN(CCOc3ccc4c5c(c(=O)oc4c3)CCCC5)CC2)c1', 'O=c1oc2cc(OCCN3CCN(Cc4ccccc4)CC3)ccc2c2c1CCCC2', 'COc1cccc(CN2CCN(CCOc3ccc4c5c(c(=O)oc4c3)CCCC5)CC2)c1', 'CN(CCCOc1ccc2c3c(c(=O)oc2c1)CCCC3)Cc1ccccc1', 'COc1cccc(CN(C)CCCOc2ccc3c4c(c(=O)oc3c2)CCCC4)c1', 'O=c1oc2cc(OCCCN3CCC(Cc4ccccc4)CC3)ccc2c2c1CCCC2', 'COc1cccc(CC2CCN(CCCOc3ccc4c5c(c(=O)oc4c3)CCCC5)CC2)c1', 'O=c1oc2cc(OCCCN3CCN(Cc4ccccc4)CC3)ccc2c2c1CCCC2', 'COc1cccc(CN2CCN(CCCOc3ccc4c5c(c(=O)oc4c3)CCCC5)CC2)c1', 'CN(CCCCOc1ccc2c3c(c(=O)oc2c1)CCCC3)Cc1ccccc1', 'COc1cccc(CN(C)CCCCOc2ccc3c4c(c(=O)oc3c2)CCCC4)c1', 'O=c1oc2cc(OCCCCN3CCC(Cc4ccccc4)CC3)ccc2c2c1CCCC2', 'COc1cccc(CC2CCN(CCCCOc3ccc4c5c(c(=O)oc4c3)CCCC5)CC2)c1', 'O=c1oc2cc(OCCCCN3CCN(Cc4ccccc4)CC3)ccc2c2c1CCCC2', 'COc1cccc(CN2CCN(CCCCOc3ccc4c5c(c(=O)oc4c3)CCCC5)CC2)c1', 'CN(CCOc1ccc2c(c1)oc(=O)c1ccccc12)Cc1ccccc1', 'COc1cccc(CN(C)CCOc2ccc3c(c2)oc(=O)c2ccccc23)c1', 'O=c1oc2cc(OCCN3CCC(Cc4ccccc4)CC3)ccc2c2ccccc12', 'COc1cccc(CC2CCN(CCOc3ccc4c(c3)oc(=O)c3ccccc34)CC2)c1', 'O=c1oc2cc(OCCN3CCN(Cc4ccccc4)CC3)ccc2c2ccccc12', 'COc1cccc(CN2CCN(CCOc3ccc4c(c3)oc(=O)c3ccccc34)CC2)c1', 'CN(CCCOc1ccc2c(c1)oc(=O)c1ccccc12)Cc1ccccc1', 'COc1cccc(CN(C)CCCOc2ccc3c(c2)oc(=O)c2ccccc23)c1', 'O=c1oc2cc(OCCCN3CCC(Cc4ccccc4)CC3)ccc2c2ccccc12', 'COc1cccc(CC2CCN(CCCOc3ccc4c(c3)oc(=O)c3ccccc34)CC2)c1', 'O=c1oc2cc(OCCCN3CCN(Cc4ccccc4)CC3)ccc2c2ccccc12', 'COc1cccc(CN2CCN(CCCOc3ccc4c(c3)oc(=O)c3ccccc34)CC2)c1', 'CN(CCCCOc1ccc2c(c1)oc(=O)c1ccccc12)Cc1ccccc1', 'COc1cccc(CN(C)CCCCOc2ccc3c(c2)oc(=O)c2ccccc23)c1', 'O=c1oc2cc(OCCCCN3CCC(Cc4ccccc4)CC3)ccc2c2ccccc12', 'COc1cccc(CC2CCN(CCCCOc3ccc4c(c3)oc(=O)c3ccccc34)CC2)c1', 'O=c1oc2cc(OCCCCN3CCN(Cc4ccccc4)CC3)ccc2c2ccccc12', 'COc1cccc(CN2CCN(CCCCOc3ccc4c(c3)oc(=O)c3ccccc34)CC2)c1', 'C[n+]1ccccc1/C=N/O.[Br-]', 'C[n+]1ccc(/C=N/O)cc1.[Br-]', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(N[C@H]1OC(=O)C3=C4CC(C)(C)C[C@H]4[C@@H](C)[C@@H](O)C[C@]31OC)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(N[C@H]1OC(=O)C3=C4CC(C)(C)C[C@H]4[C@@H](C)[C@@H](O)C[C@]31O)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(NC1OCC3=C4CC(C)(C)C[C@H]4[C@@H](C)[C@H](O)CC31O)c1ccc(=O)[nH]c1C2', 'C/C=C1\\[C@H]2C=C(C)C[C@]1(NC1OCC3=C4CC(C)(C)[C@H](O)[C@H]4[C@@H](C)CCC31O)c1ccc(=O)[nH]c1C2', 'CN(C)CCCOc1ccc(C(=O)/C=C/c2ccccc2)cc1', 'CN(C)CCCCOc1ccc(C(=O)/C=C/c2ccccc2)cc1', 'CN(C)CCCCCOc1ccc(C(=O)/C=C/c2ccccc2)cc1', 'CN(C)CCCCCCOc1ccc(C(=O)/C=C/c2ccccc2)cc1', 'CCN(CC)CCOc1ccc(C(=O)/C=C/c2ccccc2)cc1', 'CCN(CC)CCCOc1ccc(C(=O)/C=C/c2ccccc2)cc1', 'CCN(CC)CCCCOc1ccc(C(=O)/C=C/c2ccccc2)cc1', 'CCN(CC)CCCCCOc1ccc(C(=O)/C=C/c2ccccc2)cc1', 'CCN(CC)CCCCCCOc1ccc(C(=O)/C=C/c2ccccc2)cc1', 'CCCN(CCC)CCOc1ccc(C(=O)/C=C/c2ccccc2)cc1', 'CCCN(CCC)CCCOc1ccc(C(=O)/C=C/c2ccccc2)cc1', 'CCCN(CCC)CCCCOc1ccc(C(=O)/C=C/c2ccccc2)cc1', 'CCCN(CCC)CCCCCOc1ccc(C(=O)/C=C/c2ccccc2)cc1', 'CCCN(CCC)CCCCCCOc1ccc(C(=O)/C=C/c2ccccc2)cc1', 'CCCCN(CCCC)CCCCOc1ccc(C(=O)/C=C/c2ccccc2)cc1', 'CCCCN(CCCC)CCCCCOc1ccc(C(=O)/C=C/c2ccccc2)cc1', 'COc1cc2c(cc1OCCN1CCCCC1)C(=O)/C(=C/c1ccc(N(C)C)cc1)C2', 'CCN(CC)c1ccc(/C=C2\\Cc3cc(OC)c(OCCN4CCCCC4)cc3C2=O)cc1', 'CNc1ccc(/C=C2\\Cc3cc(OC)c(OCCN4CCCCC4)cc3C2=O)cc1', 'CCN(C)c1ccc(/C=C2\\Cc3cc(OC)c(OCCN4CCCCC4)cc3C2=O)cc1', 'CCCN(C)c1ccc(/C=C2\\Cc3cc(OC)c(OCCN4CCCCC4)cc3C2=O)cc1', 'CCCN(CCC)c1ccc(/C=C2\\Cc3cc(OC)c(OCCN4CCCCC4)cc3C2=O)cc1', 'CCc1nc(C(=O)NCCCCCCNc2c3c(nc4ccccc24)CCCC3)cc2c1[nH]c1ccccc12', 'Cc1nc(C(=O)NCCCCNc2c3c(nc4ccccc24)CCCC3)cc2c1[nH]c1ccccc12', 'Cc1nc(C(=O)NCCCCCCNc2c3c(nc4ccccc24)CCCC3)cc2c1[nH]c1ccccc12', 'Cc1nc(C(=O)NCCCCCCCCNc2c3c(nc4ccccc24)CCCC3)cc2c1[nH]c1ccccc12', 'O=C(NCCCCCCNc1c2c(nc3ccccc13)CCCC2)c1cc2c(cn1)[nH]c1ccccc12', 'O=C(NCCCCCCNc1c2c(nc3ccccc13)CCCC2)c1cc2c([nH]c3ccccc32)c(-c2ccccc2)n1', 'O=C(NCCCCCCNc1c2c(nc3ccccc13)CCCC2)c1cc2c([nH]c3ccccc32)c(-c2ccc(F)cc2)n1', 'COc1ccc(-c2nc(C(=O)NCCCCCCNc3c4c(nc5ccccc35)CCCC4)cc3c2[nH]c2ccccc23)cc1', 'CC(C)(C)OC(=O)N1Cc2[nH]c3ccccc3c2C[C@H]1C(=O)NCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(NCCCCCCNc1c2c(nc3ccccc13)CCCC2)[C@@H]1Cc2c([nH]c3ccccc23)CN1', 'CC1c2[nH]c3ccccc3c2C[C@@H](C(=O)NCCCCCCNc2c3c(nc4ccccc24)CCCC3)N1C(=O)OC(C)(C)C', 'CC1N[C@H](C(=O)NCCCCCCNc2c3c(nc4ccccc24)CCCC3)Cc2c1[nH]c1ccccc21', 'CC(C)(C)OC(=O)N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'CC(=O)N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'CC1=C[C@H]2Cc3nc4cc(Cl)ccc4c(NCCCCCCNc4c5c(nc6cc(Cl)ccc46)C[C@@H]4C=C(C)C[C@H]5C4)c3[C@@H](C1)C2', 'CC1=C[C@H]2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCNc4c5c(nc6cc(Cl)ccc46)C[C@@H]4C=C(C)C[C@H]5C4)c3[C@@H](C1)C2', 'CC1=C[C@H]2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCCNc4c5c(nc6cc(Cl)ccc46)C[C@@H]4C=C(C)C[C@H]5C4)c3[C@@H](C1)C2', 'CC1=C[C@H]2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCCCCNc4c5c(nc6cc(Cl)ccc46)C[C@@H]4C=C(C)C[C@H]5C4)c3[C@@H](C1)C2', 'CNc1ccccc1-c1nc(C(=O)OC)cs1', 'CNc1ccccc1-c1nc(C(N)=O)cs1', 'CNc1ccccc1-c1nc(C(=O)O)cs1', 'COc1cc(CCC(=O)/C=C/CCCNc2c3c(nc4cc(Cl)ccc24)CC2CC(C)=CC3C2)ccc1O', 'COc1cc(CCC(=O)/C=C/c2ccc(CNc3c4c(nc5cc(Cl)ccc35)CC3CC(C)=CC4C3)cc2)ccc1O', 'COc1cc(CCC(=O)/C=C/c2ccc(CNc3c4c(nc5cc(Cl)ccc35)CC3CC(C)=CC4C3)cc2)cc(CN(C)C)c1O', 'C/C=C1/[C@@H]2Cc3[nH]c(=O)ccc3[C@@]1(N)C[C@]1(C)O[C@H]21', 'COc1ccc2nc3c(c(NCCNC4=CC(=O)c5ccccc5C4=O)c2c1)CCCC3', 'COc1ccc2nc3c(c(NCCCNC4=CC(=O)c5ccccc5C4=O)c2c1)CCCC3', 'O=C1C=C(NCCNc2c3c(nc4ccccc24)CCCC3)C(=O)c2ccccc21', 'O=C1C=C(NCCCNc2c3c(nc4ccccc24)CCCC3)C(=O)c2ccccc21', 'O=C1C=C(NCCNc2c3c(nc4cc(Cl)ccc24)CCCC3)C(=O)c2ccccc21', 'O=C1C=C(NCCCNc2c3c(nc4cc(Cl)ccc24)CCCC3)C(=O)c2ccccc21', 'O=C1C(NCCNc2c3c(nc4ccccc24)CCCC3)=CC(=O)c2c(O)cccc21', 'O=C1C(NCCCNc2c3c(nc4ccccc24)CCCC3)=CC(=O)c2c(O)cccc21', 'O=C1C(NCCNc2c3c(nc4cc(Cl)ccc24)CCCC3)=CC(=O)c2c(O)cccc21', 'O=C1C(NCCNc2c3c(nc4cc(Cl)ccc24)CCCC3)=CC(=O)c2c(O)cccc21', 'O=C1C(NCCCNc2c3c(nc4cc(Cl)ccc24)CCCC3)=CC(=O)c2c(O)cccc21', 'COc1ccc2nc3c(c(NCCNC4=CC(=O)c5c(O)cccc5C4=O)c2c1)CCCC3', 'COc1ccc2nc3c(c(NCCCNC4=CC(=O)c5c(O)cccc5C4=O)c2c1)CCCC3', 'COc1ccc2nc3c(c(NCCNC4=C(Cl)C(=O)c5ccccc5C4=O)c2c1)CCCC3', 'COc1ccc2nc3c(c(NCCCNC4=C(Cl)C(=O)c5ccccc5C4=O)c2c1)CCCC3', 'O=C1C(Cl)=C(NCCNc2c3c(nc4cc(Cl)ccc24)CCCC3)C(=O)c2ccccc21', 'O=C1C(Cl)=C(NCCCNc2c3c(nc4cc(Cl)ccc24)CCCC3)C(=O)c2ccccc21', 'O=C1C(Cl)=C(NCCNc2c3c(nc4ccccc24)CCCC3)C(=O)c2ccccc21', 'O=C1C(Cl)=C(NCCCNc2c3c(nc4ccccc24)CCCC3)C(=O)c2ccccc21', 'CN1CCN(c2ccc(-c3cnc4[nH]c5cnc(C#N)cc5c4c3)cc2)CC1', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCCn1ccnc1/C=N/O)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCCn1ccnc1/C=N/O)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCCCn1ccnc1/C=N/O)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCCCn1ccnc1/C=N/O)CC2', 'COc1cc2c(cc1OC)C(c1ccccc1)N(CCCCCCn1ccnc1/C=N/O)CC2', 'COC(=O)CC1=NO[C@@H]2CCCC[C@@H]2[C@H]1c1ccccc1', 'COC(=O)CC1=NO[C@@H]2[C@H]3CC[C@H](C3)[C@@H]2[C@H]1c1ccccc1', 'O=S(=O)(CCCN1CCN(Cc2ccccc2)CC1)NCCNc1cccc2ccccc12', 'c1ccc(COC[C@H]2NC[C@H](OCc3ccccc3)[C@@H](OCc3ccccc3)[C@@H]2OCc2ccccc2)cc1', 'C=CC[C@@H]1[C@H](O)[C@@H](OCc2ccccc2)[C@H](O)CN1CCCCCCCC', 'O=C1N[C@H](COCc2ccccc2)[C@@H](OCc2ccccc2)[C@H](OCc2ccccc2)[C@H]1OCc1ccccc1', 'CO[P@]1(=O)OC[C@@H]2COC(=O)C2=C(C)O1', 'CN(CCCCCN1C(=O)c2ccccc2C1=O)Cc1ccccc1.Cl', 'CN(CCCCCCN1C(=O)c2ccccc2C1=O)Cc1ccccc1.Cl', 'CN(CCCCCCCN1C(=O)c2ccccc2C1=O)Cc1ccccc1.Cl', 'CN(CCCCCCCCN1C(=O)c2ccccc2C1=O)Cc1ccccc1.Cl', 'CN(CCCCCN1C(=O)c2ccccc2C1=O)Cc1ccccc1F.Cl', 'CN(CCCCCCN1C(=O)c2ccccc2C1=O)Cc1ccccc1F.Cl', 'CN(CCCCCCCN1C(=O)c2ccccc2C1=O)Cc1ccccc1F.Cl', 'CN(CCCCCCCCN1C(=O)c2ccccc2C1=O)Cc1ccccc1F.Cl', 'CN(CCCCCN1C(=O)c2ccccc2C1=O)Cc1cccc(Cl)c1.Cl', 'CN(CCCCCCN1C(=O)c2ccccc2C1=O)Cc1cccc(Cl)c1.Cl', 'CN(CCCCCCCN1C(=O)c2ccccc2C1=O)Cc1cccc(Cl)c1.Cl', 'CN(CCCCCCCCN1C(=O)c2ccccc2C1=O)Cc1cccc(Cl)c1.Cl', 'COc1cc2oc(-c3ccc(OCC(=O)NCCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)cc3)cc(=O)c2c(OC)c1OC', 'COc1cc2oc(-c3ccc(OCC(=O)NCCCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)cc3)cc(=O)c2c(OC)c1OC', 'COc1cc2oc(-c3cccc(OCC(=O)NCCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)c3)cc(=O)c2c(OC)c1OC', 'COc1cc2oc(-c3cccc(OCC(=O)NCCCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)c3)cc(=O)c2c(OC)c1OC', 'C#CCN(C)Cc1cc2cc(OCCCC3CCN(Cc4ccccc4C)CC3)ccc2n1C', 'C#CCN(C)Cc1cc2cc(OCCCC3CCN(Cc4ccc(OC(=O)N(C)C)c5ncccc45)CC3)ccc2n1C', 'C#CCN(C)Cc1cc2cc(OCCCC3CCN(Cc4ccccc4)CC3)ccc2[nH]1', 'C#CCNCc1cc2cc(OCCCC3CCN(Cc4ccccc4)CC3)ccc2[nH]1', 'C=CCNCc1cc2cc(OCCCC3CCN(Cc4ccccc4)CC3)ccc2[nH]1', 'CCOC(=O)c1cc2cc(OCCCC3CCN(Cc4ccccc4)CC3)ccc2[nH]1', 'OCc1cc2cc(OCCCC3CCN(Cc4ccccc4)CC3)ccc2[nH]1', 'C#CCN(C)Cc1ccc(OC(=O)N(C)C)c2ncccc12', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCNc1c3c(nc4ccccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCNc1c3c(nc4ccccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCCNc1c3c(nc4ccccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCCNc1c3c(nc4ccccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCCCNc1c3c(nc4ccccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCCCNc1c3c(nc4ccccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCCCCNc1c3c(nc4ccccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCCCCNc1c3c(nc4ccccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCCCCCNc1c3c(nc4ccccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCCCCCNc1c3c(nc4ccccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCCCCCCNc1c3c(nc4ccccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCCCCCCNc1c3c(nc4ccccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCCCCCCCCNc1c3c(nc4ccccc14)CCCC3)O2', 'O=C1c2ccccc2C(=O)N1CCCCCNCc1ccccc1', 'O=C1c2ccccc2C(=O)N1CCCCCCNCc1ccccc1', 'O=C1c2ccccc2C(=O)N1CCCCCCCNCc1ccccc1', 'O=C1c2ccccc2C(=O)N1CCCCCCCCNCc1ccccc1', 'O=C1c2ccccc2C(=O)N1CCCCCNCc1ccccc1F', 'O=C1c2ccccc2C(=O)N1CCCCCCNCc1ccccc1F', 'O=C1c2ccccc2C(=O)N1CCCCCCCNCc1ccccc1F', 'O=C1c2ccccc2C(=O)N1CCCCCCCCNCc1ccccc1F', 'O=C1c2ccccc2C(=O)N1CCCCCNCc1ccc(F)cc1', 'O=C1c2ccccc2C(=O)N1CCCCCCNCc1ccc(F)cc1', 'O=C1c2ccccc2C(=O)N1CCCCCCCNCc1ccc(F)cc1', 'O=C1c2ccccc2C(=O)N1CCCCCNCc1ccc(F)cc1F', 'O=C1c2ccccc2C(=O)N1CCCCCCNCc1ccc(F)cc1F', 'O=C1c2ccccc2C(=O)N1CCCCCCCNCc1ccc(F)cc1F', 'O=C1c2ccccc2C(=O)N1CCCCCCCCNCc1ccc(F)cc1F', 'O=C1c2ccccc2C(=O)N1CCCCCNCc1cccc(Cl)c1', 'O=C1c2ccccc2C(=O)N1CCCCCCNCc1cccc(Cl)c1', 'O=C1c2ccccc2C(=O)N1CCCCCCCNCc1cccc(Cl)c1', 'O=C1c2ccccc2C(=O)N1CCCCCCCCNCc1cccc(Cl)c1', 'Nc1oc2ccccc2c(=O)c1/C=C1\\COc2ccccc2C1=O', 'COc1cc(N)c(Cl)cc1C(=O)CCC1CCN(CC2CC2)CC1', 'COc1cc(N)c(Cl)cc1C(=O)CCC1CCN(CC2CCC2)CC1', 'COc1cc(N)c(Cl)cc1C(=O)CCC1CCN(CC2CCCC2)CC1', 'COc1cc(N)c(Cl)cc1C(=O)CCC1CCN(CC2CCCCCC2)CC1', 'COc1cc(N)c(Cl)cc1C(=O)CCC1CCN(CC2CCNCC2)CC1.Cl', 'COc1cc(N)c(Cl)cc1C(=O)CCC1CCN(CCC2CCCCC2)CC1', 'COc1cc(N)c(Cl)cc1C(=O)CCC1CCN(CC2CCCCC2C)CC1', 'COc1cc(N)c(Cl)cc1C(=O)CCC1CCN(Cc2ccccc2)CC1', 'COc1cc(N)c(Cl)cc1C(=O)CCC1CCN(Cc2cccnc2)CC1', 'COc1cc(N)c(Cl)cc1C(=O)CCC1CCN(Cc2ccncc2)CC1', 'COc1cc(N)c(Cl)cc1/C(CCC1CCN(CC2CCCCC2)CC1)=N/O', 'COc1cc(N)c(Cl)cc1C(O)CCC1CCN(CC2CCCCC2)CC1', 'COc1cc(N)c(Cl)cc1C(=O)OCC1CCN(CC2CCCCC2)CC1', 'COc1cc(N)c(Cl)cc1C(=O)CCC1CCN(CC2CCCCC2)CC1', 'COc1cc(N)c(Cl)cc1C(=O)CCC1CCN(CC2CCCCC2)CC1', 'COc1cc(N)c(Cl)cc1C(=O)CCC1CCN(CC2CCCCC2)CC1', 'COc1cc(N)c(Cl)cc1C(=O)CCC1CCN(CC2CCCCC2)CC1', 'COc1cc(N)c(Cl)cc1C(=O)CCC1CCN(CC2CCCCC2)CC1', 'Cc1c(Cl)c(=O)oc2cc(OCCN3CCN(CCCNc4c5c(nc6ccccc46)CCCC5)CC3)ccc12', 'Cc1c(C)c2ccc(OCCN3CCN(CCCNc4c5c(nc6ccccc46)CCCC5)CC3)cc2oc1=O', 'COc1cc(/C=C/c2ccc3ccccc3[n+]2C)cc(OC)c1O.[I-]', 'C[n+]1c(/C=C/c2ccccc2)ccc2ccccc21.[I-]', 'COc1cc(/C=C/c2ccc3ccccc3[n+]2C)ccc1F.[I-]', 'COc1cc(/C=C/c2ccc3ccccc3[n+]2C)cc(OC)c1OC.[I-]', 'COc1cc(/C=C/c2ccc3ccccc3[n+]2C)ccc1N1CCOCC1.[I-]', 'C[n+]1c(/C=C/c2ccc(N3CCOCC3)cc2)ccc2ccccc21.[I-]', 'C[n+]1c(/C=C/c2cc3ccccc3[nH]2)ccc2ccccc21.[I-]', 'CCn1c2ccccc2c2cc(/C=C/c3ccc4ccccc4[n+]3C)ccc21.[I-]', 'CCN(CCCOc1ccc(-c2cc(=O)c3c(OC)c(OC)c(OC)cc3o2)cc1)Cc1ccccc1OC', 'COc1cc2oc(-c3ccc(OCCCCN(C)Cc4ccccc4)cc3)cc(=O)c2c(OC)c1OC', 'CCN(CCCCOc1ccc(-c2cc(=O)c3c(OC)c(OC)c(OC)cc3o2)cc1)Cc1ccccc1', 'COc1ccccc1CN(C)CCCCOc1ccc(-c2cc(=O)c3c(OC)c(OC)c(OC)cc3o2)cc1', 'CCN(CCCCOc1ccc(-c2cc(=O)c3c(OC)c(OC)c(OC)cc3o2)cc1)Cc1ccccc1OC', 'CCN(CCCCCCOc1ccc(-c2cc(=O)c3c(OC)c(OC)c(OC)cc3o2)cc1)Cc1ccccc1OC', 'CCN(CCCOc1ccc(-c2cc(=O)c3c(O)c(OC)c(OC)cc3o2)cc1)Cc1ccccc1OC', 'CCN(CCCCOc1ccc(-c2cc(=O)c3c(O)c(OC)c(OC)cc3o2)cc1)Cc1ccccc1OC', 'CCN(CCCCCCOc1ccc(-c2cc(=O)c3c(O)c(OC)c(OC)cc3o2)cc1)Cc1ccccc1OC', 'CCC(C)C(=O)[C@]12C(=O)[C@]3(CC=C(C)C)C[C@H](CC=C(C)C)[C@@]1(C)CC[C@@H](C(C)(C)O)OC2=C(CC=C(C)C)C3=O', 'C=C(C)[C@H]1CC[C@@](C)(O)[C@@H](C2=C(O)C(CC=C(C)C)(CC=C(C)C)C(=O)C(C(=O)C(C)C)=C2O)C1', 'CC(C)=CC[C@]12C[C@@H]3[C@@H]4[C@@H](O)C(C)(C)[C@@H](C(C)C)CC4(C1=O)C(=O)[C@](C(=O)c1ccccc1)(C2=O)C3(C)C', 'C/C(=C\\C[C@]12C[C@H]3C[C@@H]4C(C)(C)CC[C@@]4(C1=O)C(=O)[C@](C(=O)c1ccccc1)(C2=O)C3(C)C)CC/C=C/C(C)C', 'CN1CCN(c2cccc(-c3cnc4[nH]c5cnc(C#N)cc5c4c3)c2)CC1', 'N#Cc1cc2c(cn1)[nH]c1ncc(-c3ccc(CN4CCCCC4)cc3)cc12', 'C[C@@H]1CCC[C@H](C)N1Cc1ccc(-c2cnc3[nH]c4cnc(C#N)cc4c3c2)cc1', 'C[C@@H]1C[C@H](C)CN(Cc2ccc(-c3cnc4[nH]c5cnc(C#N)cc5c4c3)cc2)C1', 'CC1(C)CCCN(Cc2ccc(-c3cnc4[nH]c5cnc(C#N)cc5c4c3)cc2)C1', 'CC1(C)CCN(Cc2ccc(-c3cnc4[nH]c5cnc(C#N)cc5c4c3)cc2)CC1', 'CC(C)(c1ccc(-c2cnc3[nH]c4cnc(C#N)cc4c3c2)cc1)N1CCCCC1', 'N#Cc1cc2c(cn1)[nH]c1ncc(-c3ccc(CN4CCC(O)CC4)cc3)cc12', 'N#Cc1cc2c(cn1)[nH]c1ncc(-c3ccc(CN4CCC[C@@H](O)C4)cc3)cc12', 'N#Cc1cc2c(cn1)[nH]c1ncc(-c3ccc(CN4CCC(F)CC4)cc3)cc12', 'N#Cc1cc2c(cn1)[nH]c1ncc(-c3ccc(CN4CCC(C#N)CC4)cc3)cc12', 'N#Cc1cc2c(cn1)[nH]c1ncc(-c3ccc(CN4CCOCC4)cc3)cc12', 'N#Cc1cc2c(cn1)[nH]c1ncc(-c3ccc(CN4CCC5(CC4)COC5)cc3)cc12', 'N#Cc1cc2c(cn1)[nH]c1ncc(-c3ccc(CN4CCCCC4)nc3)cc12', 'N#Cc1cc2c(cn1)[nH]c1ncc(-c3nc(CN4CCCCC4)cs3)cc12', 'CCN1CCc2cc(-c3cnc4[nH]c5cnc(C#N)cc5c4c3)ccc2C1', 'CCN1CCc2sc(-c3cnc4[nH]c5cnc(C#N)cc5c4c3)cc2C1', 'NC(=O)c1cc2c(cn1)[nH]c1ncc(-c3ccc(CN4CCCCC4)cc3)cc12', 'c1ncc(-c2cc3c(cn2)[nH]c2ncc(-c4ccc(CN5CCCCC5)cc4)cc23)o1', 'c1ncc(-c2cc3c(cn2)[nH]c2ncc(-c4ccc(CN5CCCCC5)cc4)cc23)s1', 'c1nnc(-c2cc3c(cn2)[nH]c2ncc(-c4ccc(CN5CCCCC5)cc4)cc23)o1', 'c1cc(-c2cnc3[nH]c4cnc(-c5cnoc5)cc4c3c2)ccc1CN1CCCCC1', 'Cn1cc(-c2cc3c(cn2)[nH]c2ncc(-c4ccc(CN5CCCCC5)cc4)cc23)cn1', 'c1cncc(-c2cc3c(cn2)[nH]c2ncc(-c4ccc(CN5CCCCC5)cc4)cc23)c1', 'c1cc(-c2cc3c(cn2)[nH]c2ncc(-c4ccc(CN5CCCCC5)cc4)cc23)ccn1', 'c1cnc(-c2cc3c(cn2)[nH]c2ncc(-c4ccc(CN5CCCCC5)cc4)cc23)cn1', 'c1cc(-c2cc3c(cn2)[nH]c2ncc(-c4ccc(CN5CCCCC5)cc4)cc23)cnn1', 'c1ncc(-c2cc3c(cn2)[nH]c2ncc(-c4ccc(CN5CCCCC5)cc4)cc23)cn1', 'Cc1ccc2[nH]c3c(c2c1)CN(CCCCCCN1CCc2[nH]c4ccc(C)cc4c2C1)CC3', 'Cc1ccc2[nH]c3c(c2c1)CN(CCCCCCN1CCc2[nH]c4ccc(C)cc4c2C1)CC3', 'Cc1ccc2[nH]c3c(c2c1)CN(CCCCCCCN1CCc2[nH]c4ccc(C)cc4c2C1)CC3', 'COc1ccccc1C(C)NS(=O)(=O)NC(=O)OCc1ccccc1', 'COc1ccc(C(C)NS(=O)(=O)NC(=O)OCc2ccccc2)cc1', 'COc1ccc(C(C)NS(=O)(=O)NC(=O)OCc2ccccc2)cc1OC', 'COc1ccc(OC)c(C(C)NS(=O)(=O)NC(=O)OCc2ccccc2)c1', 'COc1ccc(OC)c(C(C)NS(=O)(=O)NC(=O)OCc2ccccc2)c1', 'COc1cccc(OC)c1C(C)NS(=O)(=O)NC(=O)OCc1ccccc1', 'COc1cccc(OC)c1C(C)NS(=O)(=O)NC(=O)OCc1ccccc1', 'COc1ccccc1C(C)NS(N)(=O)=O', 'COc1ccccc1C(C)NS(N)(=O)=O', 'COc1ccc(C(C)NS(N)(=O)=O)cc1', 'COc1ccc(C(C)NS(N)(=O)=O)cc1OC', 'COc1ccc(OC)c(C(C)NS(N)(=O)=O)c1', 'COc1cccc(OC)c1C(C)NS(N)(=O)=O', 'CC1(C)C(=O)CC[C@@]2(C)[C@H]1CC[C@@]1(C)CC3=CC(=O)[C@H]4C(C)(C)[C@H](O)CC[C@]4(C)[C@H]3CC[C@@H]12', 'COc1ccc(C(=O)O[C@H]2CC[C@@]3(C)[C@@H](CC[C@@]4(C)C[C@@]5(O)[C@H](CC[C@@H]43)[C@@]3(C)CC[C@@H](O)C(C)(C)[C@@H]3C[C@@H]5O)[C@@]2(C)C(=O)O)cc1O', 'CC(=O)O[C@H]1CC[C@@]2(C)[C@@H](CC[C@@]3(C)C[C@@]4(O)CC[C@H]5C(C)(C)[C@@H](OC(C)=O)CC[C@]5(C)[C@H]4CC[C@@H]32)C1(C)C', 'CC(=O)O[C@H]1CC[C@@]2(C)[C@@H](CC[C@@]3(C)C[C@@]4(O)CC[C@H]5C(C)(C)[C@@H](OC(C)=O)CC[C@]5(C)[C@H]4CC[C@@H]32)C1(C)C', 'C[C@@]12CC[C@@H]3[C@](C)(CC[C@H](OC(=O)CCc4ccc(O)cc4)[C@]3(C)C(=O)O)[C@H]1CC[C@H]1C(=CC[C@@H]3[C@]1(C)CC[C@@H](O)[C@]3(C)CO)C2', 'C[C@@]12CC[C@@H]3[C@](C)(CC[C@H](OC(=O)CCc4ccc(O)cc4)[C@]3(C)C(=O)O)[C@H]1CC[C@H]1C(=CC[C@@H]3[C@]1(C)CC[C@@H](O)[C@]3(C)CO)C2', 'CCCC[C@@H](O)/C=C/C=C/C(=O)O', 'CCCC[C@@H](O)/C=C/C=C/C(=O)O', 'Cc1c(C)c2ccc(OCCCN(C)Cc3ccccc3)cc2oc1=O.Cl', 'Cc1c(C)c2ccc(OCCCCN(C)Cc3ccccc3)cc2oc1=O', 'Cc1c(C)c2ccc(OCCCCCNCc3ccccc3)cc2oc1=O.Cl', 'Cc1c(C)c2ccc(OCCCCCCN(C)Cc3ccccc3)cc2oc1=O', 'CCN(CCCCCCOc1ccc2c(C)c(C)c(=O)oc2c1)Cc1ccccc1.Cl', 'CN(Cc1ccccc1)Cc1ccc(COc2ccc3ccc(=O)oc3c2)cc1', 'Cc1cc2ccc(OCc3ccc(CN(C)Cc4ccccc4)cc3)cc2oc1=O.Cl', 'Cc1cc(=O)oc2cc(OCc3ccc(CN(C)Cc4ccccc4)cc3)ccc12.Cl', 'Cc1c(C)c2ccc(OCc3ccc(CN(C)Cc4ccccc4)cc3)cc2oc1=O.Cl', 'Cc1cc(=O)oc2cc(OCc3ccc(CN(C)Cc4cccc(Cl)c4)cc3)ccc12.Cl', 'CN(Cc1ccc(C#N)cc1)Cc1ccc(COc2ccc3ccc(=O)oc3c2)cc1.Cl', 'Cc1cc(=O)oc2cc(OCc3ccc(CN(C)Cc4ccc(C#N)cc4)cc3)ccc12.Cl', 'Cc1c(C)c2ccc(OCc3ccc(CN(C)Cc4ccc(C#N)cc4)cc3)cc2oc1=O.Cl', 'CN(Cc1ccccc1)Cc1ccc(COc2ccc3c(CO)cc(=O)oc3c2)cc1.Cl', 'CN(Cc1ccccc1)Cc1ccc(COc2ccc3c(CO)cc(=O)oc3c2)cc1.Cl', 'CN(Cc1ccc(COc2ccc3c(CO)cc(=O)oc3c2)cc1)Cc1cccc(Cl)c1.Cl', 'CN(Cc1ccccc1)Cc1ccc(COc2ccc3ccc(=O)[nH]c3c2)cc1.Cl', 'CN(Cc1ccccc1)Cc1ccc(COc2ccc3c(CO)cc(=O)oc3c2)cc1', 'CCOC(=O)C1=C(Nc2ccccc2)C[C@@H](c2ccccc2)N(c2ccccc2)[C@@H]1c1ccccc1', 'COC(=O)C1=C(Nc2ccccc2)C[C@@H](c2ccccc2)N(c2ccccc2)[C@@H]1c1ccccc1', 'COC(=O)C1=C(Nc2ccccc2)C[C@@H](c2ccc(C)cc2)N(c2ccccc2)[C@@H]1c1ccc(C)cc1', 'CCOC(=O)C1=C(Nc2ccccc2)C[C@@H](c2ccc(C)cc2)N(c2ccccc2)[C@@H]1c1ccc(C)cc1', 'COC(=O)C1=C(Nc2ccc(C)cc2)C[C@@H](c2ccc(C)cc2)N(c2ccc(C)cc2)[C@@H]1c1ccc(C)cc1', 'COC(=O)C1=C(Nc2ccc(Br)cc2)C[C@@H](c2ccc(C)cc2)N(c2ccc(Br)cc2)[C@@H]1c1ccc(C)cc1', 'COC(=O)C1=C(Nc2ccc(Cl)cc2)C[C@@H](c2ccc(C)cc2)N(c2ccc(Cl)cc2)[C@@H]1c1ccc(C)cc1', 'COC(=O)C1=C(Nc2ccc(OC)cc2)C[C@@H](c2ccc(C)cc2)N(c2ccc(OC)cc2)[C@@H]1c1ccc(C)cc1', 'CCOC(=O)C1=C(Nc2ccc(OC)cc2)C[C@@H](c2ccc(C)cc2)N(c2ccc(OC)cc2)[C@@H]1c1ccc(C)cc1', 'CCOC(=O)C1=C(Nc2ccc(OC)cc2)C[C@@H](c2ccccc2)N(c2ccc(OC)cc2)[C@@H]1c1ccccc1', 'CCOC(=O)C1=C(Nc2ccc(Cl)cc2)C[C@@H](c2ccccc2)N(c2ccc(Cl)cc2)[C@@H]1c1ccccc1', 'CCOC(=O)C1=C(Nc2ccc(C)cc2)C[C@@H](c2ccccc2)N(c2ccc(C)cc2)[C@@H]1c1ccccc1', 'COC(=O)C1=C(Nc2ccc(F)cc2)C[C@@H](c2ccccc2)N(c2ccc(F)cc2)[C@@H]1c1ccccc1', 'CCOC(=O)C1=C(Nc2ccc(F)cc2)C[C@@H](c2ccc(Cl)cc2)N(c2ccc(F)cc2)[C@@H]1c1ccc(Cl)cc1', 'CCOC(=O)C1=C(Nc2ccccc2)C[C@@H](c2ccc(F)cc2)N(c2ccccc2)[C@@H]1c1ccc(F)cc1', 'COC(=O)C1=C(Nc2ccc(F)cc2)C[C@@H](c2ccc(F)cc2)N(c2ccc(F)cc2)[C@@H]1c1ccc(F)cc1', 'COC(=O)C1=C(Nc2ccc(Cl)cc2)C[C@@H](c2ccc(F)cc2)N(c2ccc(Cl)cc2)[C@@H]1c1ccc(F)cc1', 'CC(C)(C)OC(=O)C1=C(Nc2ccccc2)C[C@@H](c2ccccc2)N(c2ccccc2)[C@@H]1c1ccccc1', 'COC(=O)C1=C(Nc2ccccc2)C[C@@H](c2ccc(Cl)cc2)N(c2ccccc2)[C@@H]1c1ccc(Cl)cc1', 'CCOC(=O)C1=C(Nc2ccc(C)cc2)C[C@@H](c2ccc(Cl)cc2)N(c2ccc(C)cc2)[C@@H]1c1ccc(Cl)cc1', 'COC(=O)C1=C(Nc2ccc(Br)cc2)C[C@@H](c2ccc(Cl)cc2)N(c2ccc(Br)cc2)[C@@H]1c1ccc(Cl)cc1', 'COC(=O)C1=C(Nc2ccc(Br)cc2)C[C@@H](c2ccc(OC)cc2)N(c2ccc(Br)cc2)[C@@H]1c1ccc(OC)cc1', 'CCOC(=O)C1=C(Nc2ccc(Cl)cc2)C[C@@H](c2ccc(OC)cc2)N(c2ccc(Cl)cc2)[C@@H]1c1ccc(OC)cc1', 'COC(=O)C1=C(Nc2ccccc2)C[C@@H](c2ccc3c(c2)OCO3)N(c2ccccc2)[C@@H]1c1ccc2c(c1)OCO2', 'CN(Cc1ccccc1)Cc1ccc(/C=C2\\Cc3ccc(OCCCCCN4CCC(c5ccccc5)CC4)cc3C2=O)cc1', 'CN(Cc1ccccc1)Cc1ccc(/C=C2\\CCc3cc(OCCCCCN4CCN(Cc5ccc(F)cc5)CC4)ccc3C2=O)cc1', 'O=C(C[n+]1ccc(/C=N/O)cc1)Nc1nccs1.[Cl-]', 'O=C(C[n+]1ccc(/C=N/O)cc1)Nc1nccs1.[Cl-]', 'O=C(C[n+]1ccc(/C=N/O)cc1)Nc1nccs1.[Cl-]', 'O=C(C[n+]1cccc(/C=N/O)c1)Nc1nccs1.[Cl-]', 'O=C(C[n+]1cccc(/C=N/O)c1)Nc1nccs1.[Cl-]', 'O=C(C[n+]1ccc(/C=N/O)cc1)Nc1nc(-c2ccccc2)cs1.[Cl-]', 'O=C(C[n+]1ccc(/C=N/O)cc1)Nc1nc(-c2ccccc2)cs1.[Cl-]', 'O=C(C[n+]1ccc(/C=N/O)cc1)Nc1nc(-c2ccccc2)cs1.[Cl-]', 'Cc1ccc(-c2csc(NC(=O)C[n+]3ccc(/C=N/O)cc3)n2)cc1.[Cl-]', 'Cc1ccc(-c2csc(NC(=O)C[n+]3ccc(/C=N/O)cc3)n2)cc1.[Cl-]', 'Cc1ccc(-c2csc(NC(=O)C[n+]3ccc(/C=N/O)cc3)n2)cc1.[Cl-]', 'COc1ccc(-c2csc(NC(=O)C[n+]3ccc(/C=N/O)cc3)n2)cc1.[I-]', 'COc1ccc(-c2csc(NC(=O)C[n+]3ccc(/C=N/O)cc3)n2)cc1.[I-]', 'COc1ccc(-c2csc(NC(=O)C[n+]3ccc(/C=N/O)cc3)n2)cc1.[I-]', 'CCOC(=O)Cc1csc(NC(=O)C[n+]2ccc(/C=N/O)cc2)n1.[Cl-]', 'CCOC(=O)Cc1csc(NC(=O)C[n+]2ccc(/C=N/O)cc2)n1.[Cl-]', 'CCOC(=O)Cc1csc(NC(=O)C[n+]2ccc(/C=N/O)cc2)n1.[Cl-]', 'CCOC(=O)Cc1csc(NC(=O)C[n+]2cccc(/C=N/O)c2)n1.[Cl-]', 'CCOC(=O)c1csc(NC(=O)C[n+]2ccc(/C=N/O)cc2)n1.[Cl-]', 'CCOC(=O)c1csc(NC(=O)C[n+]2ccc(/C=N/O)cc2)n1.[Cl-]', 'CCOC(=O)c1csc(NC(=O)C[n+]2ccc(/C=N/O)cc2)n1.[Cl-]', 'CCOC(=O)c1csc(NC(=O)C[n+]2cccc(/C=N/O)c2)n1.[Cl-]', 'CCOC(=O)c1csc(NC(=O)C[n+]2cccc(/C=N/O)c2)n1.[Cl-]', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCCCCNc4c5c(nc6ccccc46)CCCC5)c3C(C1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)c3C(C1)C2', 'CC1=CC2Cc3nc4ccccc4c(NCCCCCCCCCCNc4c5c(nc6ccccc46)CCCC5)c3C(C1)C2', 'CC1=CC2Cc3nc4ccccc4c(NCCCCCCCCCCCCNc4c5c(nc6ccccc46)CCCC5)c3C(C1)C2', 'CC1=CC2Cc3nc4ccccc4c(NCCCCCCCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)c3C(C1)C2', 'CC1=CC2Cc3nc4ccccc4c(NCCCCCCCCCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)c3C(C1)C2', 'CN(C)c1ccc(C(=O)NCCNc2c3c(nc4ccccc24)CCCC3)cc1.Cl', 'CN(C)c1ccc(C(=O)NCCNc2c3c(nc4ccccc24)CCCC3)cc1.Cl', 'CN(C)c1ccc(C(=O)NCCCNc2c3c(nc4ccccc24)CCCC3)cc1.Cl', 'CN(C)c1ccc(C(=O)NCCCCNc2c3c(nc4ccccc24)CCCC3)cc1.Cl', 'CN(C)c1ccc(C(=O)NCCCCCNc2c3c(nc4ccccc24)CCCC3)cc1.Cl', 'CN(C)c1ccc(C(=O)NCCCCCCNc2c3c(nc4ccccc24)CCCC3)cc1.Cl', 'CN(C)c1ccc(C(=O)NCCCCCCCNc2c3c(nc4ccccc24)CCCC3)cc1.Cl', 'CN(C)c1ccc(C(=O)NCCCCCCCCNc2c3c(nc4ccccc24)CCCC3)cc1.Cl', 'CN(C)c1ccc(C(=O)NCCCCCCCCCNc2c3c(nc4ccccc24)CCCC3)cc1.Cl', 'c1ccc2c(c1)c1cnccc1n2CCCCCCCCCn1c2ccccc2c2cnccc21', 'C[n+]1ccc2c(c1)c1ccccc1n2CCCCCCCCCn1c2ccccc2c2c[n+](C)ccc21.[I-].[I-]', 'Cc1ccccc1-c1nnc(-c2ccc3c(c2)CO[C@@]3(CCCN(C)C)c2ccc(F)cc2)[nH]1', 'CN(C)CCC[C@@]1(c2ccc(F)cc2)OCc2cc(-c3nnc(-c4ccc(Cl)cc4)[nH]3)ccc21', 'CC[N+](CC)(CCCCCn1c(C)cc(=O)n(CCCCC[N+](CC)(CC)Cc2ccccc2[N+](=O)[O-])c1=O)Cc1ccccc1[N+](=O)[O-].[Br-].[Br-]', 'CC[N+](CC)(CCCCCn1c(C)cc(=O)n(CCCCC[N+](CC)(CC)Cc2ccccc2[N+](=O)[O-])c1=O)Cc1ccccc1[N+](=O)[O-].[Br-].[Br-]', 'CC[N+]1(Cc2ccccc2[N+](=O)[O-])CCCCCC[N+](CC)(Cc2ccccc2[N+](=O)[O-])CCCCCn2c(=O)cc(C)n(c2=O)CCCCC1.[Br-].[Br-]', 'CC[N+]1(Cc2ccccc2[N+](=O)[O-])CCCCCn2c(C)cc(=O)n(c2=O)CCCCC[N+](CC)(Cc2ccccc2[N+](=O)[O-])Cc2ccc(cc2)C1.[Br-].[Br-]', 'CC[N+]1(Cc2ccccc2[N+](=O)[O-])CCCCCn2c(C)cc(=O)n(c2=O)CCCCC[N+](CC)(Cc2ccccc2[N+](=O)[O-])Cc2ccc(cc2)C1.[Br-].[Br-]', 'CC[N+]1(Cc2ccccc2[N+](=O)[O-])CCCCCn2c(C)cc(=O)n(c2=O)CCCCC[N+](CC)(Cc2ccccc2[N+](=O)[O-])CCCCCn2c(=O)cc(C)n(c2=O)CCCCC1.[Br-].[Br-]', 'CC[N+]1(Cc2ccccc2[N+](=O)[O-])CCCCCn2c(C)c3c(=O)n(c2=O)CCCCC[N+](CC)(Cc2ccccc2[N+](=O)[O-])CCCCCn2c(C)c(c(=O)n(c2=O)CCCCC1)C3.[Br-].[Br-]', 'COc1ccc2cc3[n+](cc2c1OCc1ccccc1)CCc1cc2c(cc1-3)OCO2.[Br-]', 'CC(C)n1cc(C(=O)C(F)(F)F)cn1', 'CC(C)n1cc(C(=O)C(F)(F)F)cn1', 'CCC(C)n1cc(C(=O)C(F)(F)F)cn1', 'CCC(C)n1cc(C(=O)C(F)(F)F)cn1', 'CCCC(C)n1cc(C(=O)C(F)(F)F)cn1', 'CCCC(C)n1cc(C(=O)C(F)(F)F)cn1', 'CCC(CC)n1cc(C(=O)C(F)(F)F)cn1', 'CCC(CC)n1cc(C(=O)C(F)(F)F)cn1', 'O=C(c1cnn(C2CCCC2)c1)C(F)(F)F', 'O=C(c1cnn(C2CCCC2)c1)C(F)(F)F', 'CC(C)Cn1cc(C(=O)C(F)(F)F)cn1', 'CC(C)Cn1cc(C(=O)C(F)(F)F)cn1', 'CC(C)(C)c1cccc(C(=O)C(F)F)c1', 'CC(C)(C)c1cccc(C(=O)C(F)F)c1', 'CC(C)n1cc(C(=O)C(F)F)cn1', 'CC(C)n1cc(C(=O)C(F)F)cn1', 'CCC(C)n1cc(C(=O)C(F)F)cn1', 'CCC(C)n1cc(C(=O)C(F)F)cn1', 'CCCC(C)n1cc(C(=O)C(F)F)cn1', 'CCCC(C)n1cc(C(=O)C(F)F)cn1', 'CCC(CC)n1cc(C(=O)C(F)F)cn1', 'CCC(CC)n1cc(C(=O)C(F)F)cn1', 'O=C(c1cnn(C2CCCC2)c1)C(F)F', 'O=C(c1cnn(C2CCCC2)c1)C(F)F', 'CC(C)Cn1cc(C(=O)C(F)F)cn1', 'CC(C)Cn1cc(C(=O)C(F)F)cn1', 'CC(C)(C)c1cccc(C(=O)CF)c1', 'CC(C)(C)c1cccc(C(=O)CF)c1', 'CCC(C)n1cc(C(=O)CF)cn1', 'CCC(C)n1cc(C(=O)CF)cn1', 'CCCC(C)n1cc(C(=O)CF)cn1', 'CCCC(C)n1cc(C(=O)CF)cn1', 'CCC(CC)n1cc(C(=O)CF)cn1', 'CCC(CC)n1cc(C(=O)CF)cn1', 'O=C(CF)c1cnn(C2CCCC2)c1', 'O=C(CF)c1cnn(C2CCCC2)c1', 'O=C(NCCc1c[nH]c2ccccc12)c1ccc[n+](Cc2ccccc2)c1.[Br-]', 'O=C(NCCc1c[nH]c2ccccc12)c1ccc[n+](Cc2ccccc2F)c1.[Br-]', 'O=C(NCCc1c[nH]c2ccccc12)c1cc[n+](Cc2ccccc2)cc1.[Br-]', 'O=C(NCCc1c[nH]c2ccccc12)c1cc[n+](Cc2ccccc2F)cc1.[Br-]', 'O=C(NCCc1c[nH]c2ccccc12)c1cc[n+](Cc2cccc(F)c2)cc1.[Br-]', 'O=C(NCCc1c[nH]c2ccccc12)c1cc[n+](Cc2ccc(F)cc2)cc1.[Br-]', 'Cc1cccc(C[n+]2ccc(C(=O)NCCc3c[nH]c4ccccc34)cc2)c1.[Br-]', 'Cc1ccc(C[n+]2ccc(C(=O)NCCc3c[nH]c4ccccc34)cc2)cc1.[Br-]', 'N#Cc1ccc(C[n+]2ccc(C(=O)NCCc3c[nH]c4ccccc34)cc2)cc1.[Br-]', 'COc1ccc2[nH]cc(CCNC(=O)c3cc[n+](Cc4ccccc4F)cc3)c2c1.[Br-]', 'COc1ccc2[nH]cc(CCNC(=O)c3cc[n+](Cc4cccc(F)c4)cc3)c2c1.[Br-]', 'COc1ccc2[nH]cc(CCNC(=O)c3cc[n+](Cc4ccc(F)cc4)cc3)c2c1.[Br-]', 'COc1ccc2[nH]cc(CCNC(=O)c3cc[n+](Cc4ccc(OC(=O)N(C)C)cc4)cc3)c2c1.[Br-]', 'COc1ccc2[nH]cc(CCNC(=O)/C=C/c3cc[n+](Cc4ccccc4F)cc3)c2c1.[Br-]', 'COc1ccc2[nH]cc(CCNC(=O)/C=C/c3cc[n+](Cc4cccc(F)c4)cc3)c2c1.[Br-]', 'COc1ccc2[nH]cc(CCNC(=O)/C=C/c3cc[n+](Cc4ccc(F)cc4)cc3)c2c1.[Br-]', 'COc1ccc2[nH]cc(CCNC(=O)CCc3cc[n+](Cc4ccccc4F)cc3)c2c1.[Br-]', 'COc1ccc2[nH]cc(CCNC(=O)CCc3cc[n+](Cc4cccc(F)c4)cc3)c2c1.[Br-]', 'COc1ccc2[nH]cc(CCNC(=O)CCc3cc[n+](Cc4ccc(F)cc4)cc3)c2c1.[Br-]', 'COc1ccc2nc3c(c(NCCNC(=O)C4(C)CCc5c(C)c(O)c(C)c(C)c5O4)c2c1)CCCC3', 'COc1ccc2nc3c(c(NCCCNC(=O)C4(C)CCc5c(C)c(O)c(C)c(C)c5O4)c2c1)CCCC3', 'COc1ccc2nc3c(c(NCCCCNC(=O)C4(C)CCc5c(C)c(O)c(C)c(C)c5O4)c2c1)CCCC3', 'COc1ccc2nc3c(c(NCCCCCNC(=O)C4(C)CCc5c(C)c(O)c(C)c(C)c5O4)c2c1)CCCC3', 'COc1ccc2nc3c(c(NCCCCCCNC(=O)C4(C)CCc5c(C)c(O)c(C)c(C)c5O4)c2c1)CCCC3', 'COc1ccc2nc3c(c(NCCCCCCCNC(=O)C4(C)CCc5c(C)c(O)c(C)c(C)c5O4)c2c1)CCCC3', 'COc1ccc2nc3c(c(NCCCCCCCCNC(=O)C4(C)CCc5c(C)c(O)c(C)c(C)c5O4)c2c1)CCCC3', 'COc1ccc(-c2cc3c(c(=O)o2)C[C@H]2[C@@](C)(CC[C@@]4(O)C(C)(C)C(=O)CC[C@]24C)O3)cc1', 'COc1ccc(-c2cc3c(c(=O)o2)C[C@]2(O)[C@@]4(C)CC[C@@H](OC(C)=O)C(C)(C)[C@]4(O)CC[C@@]2(C)O3)cc1', 'COc1cc(-c2cc3c(c(=O)o2)C[C@]2(O)[C@@]4(C)C(=O)C=CC(C)(C)[C@]4(O)CC[C@@]2(C)O3)cc(OC)c1O', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCNc1c3c(nc4ccccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'Cc1c(C)c2c(c(C)c1O)CCC(C)(C(=O)NCCCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'CC(=O)NCCc1c[nH]c2ccc(OC(=O)NCCCCCCCNc3c4c(nc5ccccc35)CCCC4)cc12', 'CC(=O)NCCc1c[nH]c2ccc(OC(=O)NCCNc3c4c(nc5ccccc35)CCCC4)cc12', 'CC(=O)NCCc1c[nH]c2ccc(OC(=O)NCCNc3c4c(nc5ccccc35)CCCC4)cc12', 'CC(=O)NCCc1c[nH]c2ccc(OC(=O)NCCCCCCNc3c4c(nc5ccccc35)CCCC4)cc12', 'CC(=O)NCCc1c[nH]c2ccc(OC(=O)NCCCCCCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)cc12', 'CC(=O)NCCc1c[nH]c2ccc(OC(=O)NCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)cc12', 'CC(=O)O[C@@]12CCN(C(C)=O)[C@@H]1N(C(C)=O)c1ccc(OC(=O)NCCCCCCNc3c4c(nc5ccccc35)CCCC4)cc12', 'CC(=O)NCCc1c[nH]c2ccc(OC(=O)NCCCNc3c4c(nc5ccccc35)CCCC4)cc12', 'CC(=O)NCCc1c[nH]c2ccc(OC(=O)NCCCCNc3c4c(nc5ccccc35)CCCC4)cc12', 'CC(=O)NCCc1c[nH]c2ccc(OC(=O)NCCCCCNc3c4c(nc5ccccc35)CCCC4)cc12', 'CC(=O)NCCc1c[nH]c2ccc(OC(=O)NCCCCCCCCNc3c4c(nc5ccccc35)CCCC4)cc12', 'CC(=O)NCCc1c[nH]c2ccc(OC(=O)NCCCCCCCNc3c4c(nc5ccc(Cl)cc35)CCCC4)cc12', 'CC(=O)NCCc1c[nH]c2ccc(OC(=O)NCCNc3c4c(nc5ccc(Cl)cc35)CCCC4)cc12', 'CC(=O)NCCc1c[nH]c2ccc(OC(=O)NCCCCCCCCCCNc3c4c(nc5ccccc35)CCCC4)cc12', 'CC(=O)NCCc1c[nH]c2ccc(OC(=O)NCCCCCCCCCCCCNc3c4c(nc5ccccc35)CCCC4)cc12', 'CC(=O)NCCc1c[nH]c2ccc(OC(=O)NCCCCCCCCCCNc3c4c(nc5ccc(Cl)cc35)CCCC4)cc12', 'CC(=O)NCCc1c[nH]c2ccc(OC(=O)NCCCCCCCCCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)cc12', 'CC(=O)NCCc1c[nH]c2ccc(OC(=O)NCCCCCCCCCCCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)cc12', 'CC(=O)NCCc1c[nH]c2ccc(OC(=O)NCCCCCCCCCCCCNc3c4c(nc5ccc(Cl)cc35)CCCC4)cc12', 'Cc1c([C@@H](O)CN2CCN(C[C@H](O)c3ccc4c(c3C)COC4=O)CC2)ccc2c1COC2=O', 'CCOP1(=O)CCCC(C(=O)OC)=C(C)O1', 'CCOP1(=O)CCCC(C(=O)OC)=C(C)O1', 'COP1(=O)OCCC(C(=O)OC(C)(C)C)=C(C)O1', 'COP1(=O)OCCC(C(=O)OC(C)(C)C)=C(C)O1', 'COC(=O)C1=C(C)OP(=O)(OC)OCC1', 'COC(=O)C1=C(C)OP(=O)(OC)OCC1', 'O=C1OCc2cc(-c3ccc(-c4ccc(CO)cc4)s3)ccc21', 'CCN(CCCCCCCCc1cccc(OC)c1)Cc1ccccc1OC', 'CCN(CCCCCCCCc1cccc(OC)c1)Cc1ccccc1OC', 'O=C(NCCCCCCNc1c2c(nc3ccccc13)CCC2)c1cc2ccccc2o1', 'O=C(NCCCCCCNc1c2c(nc3ccccc13)CCCC2)c1cc2ccccc2o1', 'O=C(NCCCCCCNc1c2c(nc3ccccc13)CCCCC2)c1cc2ccccc2o1', 'COc1cccc2cc(C(=O)NCCCCCCNc3c4c(nc5ccccc35)CCC4)oc12', 'COc1cccc2cc(C(=O)NCCCCCCNc3c4c(nc5ccccc35)CCCC4)oc12', 'COc1cccc2cc(C(=O)NCCCCCCNc3c4c(nc5ccccc35)CCCCC4)oc12', 'O=C(NCCCCCCCNc1c2c(nc3ccccc13)CCC2)c1cc2ccccc2o1', 'O=C(NCCCCCCCNc1c2c(nc3ccccc13)CCCC2)c1cc2ccccc2o1', 'O=C(NCCCCCCCNc1c2c(nc3ccccc13)CCCCC2)c1cc2ccccc2o1', 'COc1cccc2cc(C(=O)NCCCCCCCNc3c4c(nc5ccccc35)CCC4)oc12', 'COc1cccc2cc(C(=O)NCCCCCCCNc3c4c(nc5ccccc35)CCCC4)oc12', 'COc1cccc2cc(C(=O)NCCCCCCCNc3c4c(nc5ccccc35)CCCCC4)oc12', 'c1ccc2oc(CNCCCCCCNc3c4c(nc5ccccc35)CCC4)cc2c1', 'COc1cccc2cc(CNCCCCCCNc3c4c(nc5ccccc35)CCC4)oc12', 'COc1cccc2cc(CNCCCCCCNc3c4c(nc5ccccc35)CCCC4)oc12', 'c1ccc2oc(CNCCCCCCCNc3c4c(nc5ccccc35)CCC4)cc2c1', 'c1ccc2oc(CNCCCCCCCNc3c4c(nc5ccccc35)CCCC4)cc2c1', 'c1ccc2oc(CNCCCCCCCNc3c4c(nc5ccccc35)CCCC4)cc2c1', 'c1ccc2oc(CNCCCCCCCNc3c4c(nc5ccccc35)CCCC4)cc2c1', 'COc1cccc2cc(CNCCCCCCCNc3c4c(nc5ccccc35)CCCCC4)oc12', 'c1ccc2oc(CN(CCCCCCNc3c4c(nc5ccccc35)CCC4)Cc3cc4ccccc4o3)cc2c1', 'COc1cccc2cc(CN(CCCCCCNc3c4c(nc5ccccc35)CCC4)Cc3cc4cccc(OC)c4o3)oc12', 'COc1cccc2cc(CN(CCCCCCNc3c4c(nc5ccccc35)CCCC4)Cc3cc4cccc(OC)c4o3)oc12', 'COc1cccc2cc(CN(CCCCCCNc3c4c(nc5ccccc35)CCCCC4)Cc3cc4cccc(OC)c4o3)oc12', 'c1ccc2oc(CN(CCCCCCCNc3c4c(nc5ccccc35)CCC4)Cc3cc4ccccc4o3)cc2c1', 'c1ccc2oc(CN(CCCCCCCNc3c4c(nc5ccccc35)CCCC4)Cc3cc4ccccc4o3)cc2c1', 'c1ccc2oc(CN(CCCCCCCNc3c4c(nc5ccccc35)CCCCC4)Cc3cc4ccccc4o3)cc2c1', 'COc1cccc2cc(CN(CCCCCCCNc3c4c(nc5ccccc35)CCCCC4)Cc3cc4cccc(OC)c4o3)oc12', 'Cl.Cl.c1ccc2c(NCCCCCCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'CC(C)NCc1cccc2[nH]c3ccccc3c12', 'CCNCc1cccc2[nH]c3cc(Cl)ccc3c12', 'COc1cc(/C=C/C(=O)N(Cc2cccc3[nH]c4ccccc4c23)C(C)C)ccc1O', 'CCN(Cc1cccc2[nH]c3cc(Cl)ccc3c12)C(=O)/C=C/c1ccc(O)c(OC)c1', 'I.ICC1CN=C(Nc2ccccn2)S1', 'CC(C)=CC[C@H]1C[C@@]2(CC=C(C)C)C(=O)O[C@@](CC=C(C)C)(C(=O)[C@@H](C(=O)C(C)C)[C@]1(C)C/C=C/C(C)(C)O)C2=O', 'CCC(C)C(=O)[C@@H]1C(=O)[C@]2(CC=C(C)C)OC(=O)[C@](CC=C(C)C)(C[C@H](CC=C(C)C)[C@@]1(C)C/C=C/C(C)(C)O)C2=O', 'NS(=O)(=O)c1ccc(/N=C/c2cc(Br)cc(Br)c2O)cc1', 'NS(=O)(=O)c1ccc(/N=C/c2cc(Br)cc(Br)c2O)cc1', 'CC[C@H]1O[C@@H]2O[C@H](/C=C/C=C/C=C/c3oc(=O)cc(OC/C=C(\\C)C(=O)OC)c3C)[C@H](O)[C@]2(C)[C@@]1(C)O', 'COc1ccc2c3c1O[C@H]1CC(O)CC[C@@]31CCN(C)C2', 'N#CC1=C(N)N2CCNC2=C([N+](=O)[O-])C12C(=O)Nc1ccccc12', 'N#CC1=C(N)N2CCCNC2=C([N+](=O)[O-])C12C(=O)Nc1ccccc12', 'CC1(C)CNC2=C([N+](=O)[O-])C3(C(=O)Nc4ccccc43)C(C#N)=C(N)N2C1', 'N#CC1=C(N)N(Cc2ccccc2)C(NCc2ccccc2)=C([N+](=O)[O-])C12C(=O)Nc1ccccc12', 'Cc1ccc2c(c1)C1(C(=O)N2)C(C#N)=C(N)N2CCNC2=C1[N+](=O)[O-]', 'Cc1ccc2c(c1)C1(C(=O)N2)C(C#N)=C(N)N2CCCNC2=C1[N+](=O)[O-]', 'Cc1ccc2c(c1)C1(C(=O)N2)C(C#N)=C(N)N2CC(C)(C)CNC2=C1[N+](=O)[O-]', 'Cc1ccc2c(c1)C1(C(=O)N2)C(C#N)=C(N)N(Cc2ccccc2)C(NCc2ccccc2)=C1[N+](=O)[O-]', 'N#CC1=C(N)N2CCNC2=C([N+](=O)[O-])C12C(=O)Nc1ccc([N+](=O)[O-])cc12', 'N#CC1=C(N)N2CCCNC2=C([N+](=O)[O-])C12C(=O)Nc1ccc([N+](=O)[O-])cc12', 'CC1(C)CNC2=C([N+](=O)[O-])C3(C(=O)Nc4ccc([N+](=O)[O-])cc43)C(C#N)=C(N)N2C1', 'N#CC1=C(N)N(Cc2ccccc2)C(NCc2ccccc2)=C([N+](=O)[O-])C12C(=O)Nc1ccc([N+](=O)[O-])cc12', 'N#CC1=C(N)N2CCNC2=C([N+](=O)[O-])C12C(=O)Nc1ccc(F)cc12', 'N#CC1=C(N)N2CCCNC2=C([N+](=O)[O-])C12C(=O)Nc1ccc(F)cc12', 'CC1(C)CNC2=C([N+](=O)[O-])C3(C(=O)Nc4ccc(F)cc43)C(C#N)=C(N)N2C1', 'N#CC1=C(N)N(Cc2ccccc2)C(NCc2ccccc2)=C([N+](=O)[O-])C12C(=O)Nc1ccc(F)cc12', 'N#CC1=C(N)N2CCNC2=C([N+](=O)[O-])C12C(=O)c1cccc3cccc2c13', 'N#CC1=C(N)N2CCCNC2=C([N+](=O)[O-])C12C(=O)c1cccc3cccc2c13', 'CC1(C)CNC2=C([N+](=O)[O-])C3(C(=O)c4cccc5cccc3c45)C(C#N)=C(N)N2C1', 'N#CC1=C(N)N(Cc2ccccc2)C(NCc2ccccc2)=C([N+](=O)[O-])C12C(=O)c1cccc3cccc2c13', 'N#CC1=C(N)N2CCNC2=C([N+](=O)[O-])C12C(=O)c1ccc(Br)c3c(Br)ccc2c13', 'N#CC1=C(N)N2CCCNC2=C([N+](=O)[O-])C12C(=O)c1ccc(Br)c3c(Br)ccc2c13', 'CC1(C)CNC2=C([N+](=O)[O-])C3(C(=O)c4ccc(Br)c5c(Br)ccc3c45)C(C#N)=C(N)N2C1', 'N#CC1=C(N)N(Cc2ccccc2)C(NCc2ccccc2)=C([N+](=O)[O-])C12C(=O)c1ccc(Br)c3c(Br)ccc2c13', 'O=c1cc(Cl)c2ccc(OCCNC3CCN(Cc4ccccc4)CC3)cc2o1', 'Cc1c(C)c2ccc(OCCNC3CCN(Cc4ccccc4)CC3)cc2oc1=O', 'Cc1c(Cl)c(=O)oc2cc(OCCNC3CCN(Cc4ccccc4)CC3)ccc12', 'O=c1oc2cc(OCCNC3CCN(Cc4ccccc4)CC3)ccc2c2c1CCCC2', 'Cc1cc(=O)oc2cc(CNCCC3CCN(Cc4ccccc4)CC3)ccc12', 'COc1ccc(C(=O)N2c3ccccc3Sc3ccccc32)cc1OC', 'O=C(Cc1ccccc1)N1c2ccccc2Sc2ccc(Cl)cc21', 'O=C(CCc1ccccc1)N1c2ccccc2Sc2ccc(Cl)cc21', 'COc1cccc(C(=O)N2c3ccccc3Sc3ccc(Cl)cc32)c1', 'COc1ccc(C(=O)N2c3ccccc3Sc3ccc(Cl)cc32)cc1OC', 'c1ccc2c(c1)Nc1ccccc1[Se]2', 'Clc1ccc2c(c1)Nc1ccccc1[Se]2', 'O=C(c1ccccc1)N1c2ccccc2[Se]c2ccccc21', 'O=C(Cc1ccccc1)N1c2ccccc2[Se]c2ccccc21', 'O=C(CCc1ccccc1)N1c2ccccc2[Se]c2ccccc21', 'COc1ccc(C(=O)N2c3ccccc3[Se]c3ccccc32)cc1', 'COc1cccc(C(=O)N2c3ccccc3[Se]c3ccccc32)c1', 'COc1ccc(C(=O)N2c3ccccc3[Se]c3ccccc32)cc1OC', 'O=C(c1ccccc1)N1c2ccccc2[Se]c2ccc(Cl)cc21', 'O=C(Cc1ccccc1)N1c2ccccc2[Se]c2ccc(Cl)cc21', 'O=C(CCc1ccccc1)N1c2ccccc2[Se]c2ccc(Cl)cc21', 'COc1ccc(C(=O)N2c3ccccc3[Se]c3ccc(Cl)cc32)cc1', 'COc1cccc(C(=O)N2c3ccccc3[Se]c3ccc(Cl)cc32)c1', 'COc1ccc(C(=O)N2c3ccccc3[Se]c3ccc(Cl)cc32)cc1OC', 'CC(COC(=O)c1ccc(Oc2nc(Oc3ccc(C(=O)OCC(C)N(C)C)cc3)nc(Oc3ccc(C(=O)OCC(C)N(C)C)cc3)n2)cc1)N(C)C', 'C[N+](C)(C)CCOC(=O)c1ccc(Oc2nc(Oc3ccc(C(=O)OCC[N+](C)(C)C)cc3)nc(Oc3ccc(C(=O)OCC[N+](C)(C)C)cc3)n2)cc1', 'CCCCCCCNC(=O)Oc1ccc2c(c1)CN(C)C(c1ccc(F)cc1)N2C', 'CCCCCCCNC(=O)Oc1ccc2c(c1)CN(C)C(c1cc[nH]c1)N2C', 'CN1Cc2cc(O)ccc2N(C)C1c1c(Cl)cccc1Cl', 'COc1cccc2c1CC[C@H](NC(=O)N(C)C)C2', 'COc1cccc2c1CC[C@H](NC(=O)N(C)C)C2', 'COc1ccc2c(c1)CCC(NC(=O)N(C)C)C2', 'COc1ccc2c(c1)CCC(NC(=O)N(C)C)C2', 'COc1ccc2c(c1)C[C@@H](NC(=O)N(C)C)CC2', 'COc1ccc2c(c1)C[C@@H](NC(=O)N(C)C)CC2', 'COc1cccc2c1CCC(NS(=O)(=O)N(C)C)C2', 'COc1cccc2c1CCC(NS(=O)(=O)N(C)C)C2', 'COc1ccc2c(c1)CCC(NS(=O)(=O)N(C)C)C2', 'COc1ccc2c(c1)CCC(NS(=O)(=O)N(C)C)C2', 'COc1ccc2c(c1)CC(NS(=O)(=O)N(C)C)CC2', 'COc1ccc2c(c1)CC(NS(=O)(=O)N(C)C)CC2', 'CN(C)C(=O)N[C@H]1CCc2c(O)cccc2C1', 'CN(C)C(=O)N[C@H]1CCc2c(O)cccc2C1', 'CN(C)C(=O)NC1CCc2cc(O)ccc2C1', 'CN(C)C(=O)NC1CCc2cc(O)ccc2C1', 'CN(C)C(=O)N[C@H]1CCc2ccc(O)cc2C1', 'CN(C)C(=O)N[C@H]1CCc2ccc(O)cc2C1', 'CN(C)S(=O)(=O)NC1CCc2c(O)cccc2C1', 'CN(C)S(=O)(=O)NC1CCc2c(O)cccc2C1', 'CN(C)S(=O)(=O)N[C@H]1CCc2cc(O)ccc2C1', 'CN(C)S(=O)(=O)N[C@H]1CCc2cc(O)ccc2C1', 'CN(C)S(=O)(=O)NC1CCc2ccc(O)cc2C1', 'CN(C)S(=O)(=O)NC1CCc2ccc(O)cc2C1', 'COc1cc2occ(-c3ccc(O)cc3O)c(=O)c2c(O)c1CC=C(C)C', 'COc1cc(/C=C/C(=O)NCCC2CCN(Cc3ccccc3)CC2)ccc1O', 'COc1cc(/C=C/C(=O)NCCC2CCN(Cc3ccccc3)CC2)ccc1O', 'COc1ccc(/C=C/C(=O)NCCC2CCN(Cc3ccccc3)CC2)cc1O', 'COc1cc(/C=C/C(=O)NCCC2CCN(Cc3ccccc3)CC2)cc(OC)c1O', 'O=C(/C=C/c1ccc(O)cc1)NCCC1CCN(Cc2ccccc2)CC1', 'O=C(/C=C/c1ccc(O)cc1)NCCC1CCN(Cc2ccccc2)CC1', 'O=C(/C=C/c1ccc(O)cc1)NCCC1CCN(Cc2ccccc2)CC1', 'O=C(/C=C/c1ccc(O)c(O)c1)NCCC1CCN(Cc2ccccc2)CC1', 'O=C(/C=C/c1ccc(O)c(O)c1)NCCC1CCN(Cc2ccccc2)CC1', 'O=C(/C=C/c1ccccc1O)NCCC1CCN(Cc2ccccc2)CC1', 'O=C(/C=C/c1ccccc1)NCCC1CCN(Cc2ccccc2)CC1', 'O=C(/C=C/c1ccccc1)NCCC1CCN(Cc2ccccc2)CC1', 'COc1cc(/C=C/C(=O)NCCC2CCN(Cc3ccccc3)CC2)ccc1OC(C)=O', 'COc1ccc(/C=C/C(=O)NCCC2CCN(Cc3ccccc3)CC2)cc1OC', 'COc1ccc(/C=C/C(=O)NCCC2CCN(Cc3ccccc3)CC2)cc1OC', 'COc1cccc(/C=C/C(=O)NCCC2CCN(Cc3ccccc3)CC2)c1', 'COc1cccc(/C=C/C(=O)NCCC2CCN(Cc3ccccc3)CC2)c1', 'COc1cc(/C=C/C(=O)NCCC2CCN(Cc3ccccc3)CC2)cc(OC)c1OC', 'O=C(/C=C/c1ccc(F)cc1)NCCC1CCN(Cc2ccccc2)CC1', 'O=C(/C=C/c1ccc(Cl)cc1)NCCC1CCN(Cc2ccccc2)CC1', 'O=C(/C=C/c1ccc([N+](=O)[O-])cc1)NCCC1CCN(Cc2ccccc2)CC1', 'COc1cc(CCC(=O)NCCC2CCN(Cc3ccccc3)CC2)ccc1O', 'COc1cc2c(cc1O)C(=O)/C(=C/C1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1O)C(=O)/C(=C/C1CCN(Cc3ccccc3)CC1)C2', 'COc1cc2c(cc1O)C/C(=C\\C1CCN(Cc3ccccc3)CC1)C2=O', 'COc1cc2c(cc1O)C/C(=C\\C1CCN(Cc3ccccc3)CC1)C2=O', 'O=C1/C(=C/C2CCN(Cc3ccccc3)CC2)Cc2ccc(O)cc21', 'O=C1/C(=C/C2CCN(Cc3ccccc3)CC2)Cc2ccc(O)cc21', 'O=C1/C(=C/C2CCN(Cc3ccccc3)CC2)Cc2cccc(O)c21', 'O=C1/C(=C/C2CCN(Cc3ccccc3)CC2)Cc2cccc(O)c21', 'O=C1/C(=C/C2CCN(Cc3ccccc3)CC2)Cc2cc(O)ccc21', 'O=C1/C(=C/C2CCN(Cc3ccccc3)CC2)Cc2cc(O)ccc21', 'COc1ccc2c(c1)C(=O)/C(=C/C1CCN(Cc3ccccc3)CC1)C2', 'COc1ccc2c(c1)C(=O)/C(=C/C1CCN(Cc3ccccc3)CC1)C2', 'COc1cccc2c1C(=O)/C(=C/C1CCN(Cc3ccccc3)CC1)C2', 'COc1cccc2c1C(=O)/C(=C/C1CCN(Cc3ccccc3)CC1)C2', 'COc1ccc2c(c1)C/C(=C\\C1CCN(Cc3ccccc3)CC1)C2=O', 'COc1ccc2c(c1)C/C(=C\\C1CCN(Cc3ccccc3)CC1)C2=O', 'O=C1/C(=C/C2CCN(Cc3ccccc3)CC2)Cc2cc3c(cc21)OCO3', 'O=C1/C(=C/C2CCN(Cc3ccccc3)CC2)Cc2cc3c(cc21)OCO3', 'COc1cccc(CN2CCC(/C=C3\\Cc4cc(OC)ccc4C3=O)CC2)c1', 'COc1cccc(CN2CCC(/C=C3\\Cc4cc(OC)ccc4C3=O)CC2)c1', 'COc1cccc(CN2CCC(/C=C3\\Cc4ccc(OC)cc4C3=O)CC2)c1', 'COc1cccc(CN2CCC(/C=C3\\Cc4cccc(OC)c4C3=O)CC2)c1', 'COc1cccc(CN2CCC(/C=C3\\Cc4cccc(OC)c4C3=O)CC2)c1', 'COc1ccc2c(c1)C(=O)/C(=C/C1CCN(Cc3ccc4c(c3)OCO4)CC1)C2', 'COc1ccc2c(c1)C(=O)/C(=C/C1CCN(Cc3ccc4c(c3)OCO4)CC1)C2', 'COc1cccc2c1C(=O)/C(=C/C1CCN(Cc3ccc4c(c3)OCO4)CC1)C2', 'COc1cccc2c1C(=O)/C(=C/C1CCN(Cc3ccc4c(c3)OCO4)CC1)C2', 'O=C(CCCCCCCCc1ccc(O)cc1)c1c(O)cc(O)cc1O', 'O=C(CCCCCCCCc1ccc(O)c(O)c1)c1c(O)ccc(C(CCCCCCCCc2ccc(O)c(O)c2)c2c(O)cccc2O)c1O', 'O=C(CCCCCCCCc1ccc(O)c(O)c1)c1c(O)ccc(C(CCCCCCCCc2ccc(O)cc2)c2c(O)cccc2O)c1O', 'O=C(CC1CCN(Cc2ccccc2)CC1)NCCc1c[nH]c2ccccc12', 'Cc1ccccc1CN1CCC(CC(=O)NCCc2c[nH]c3ccccc23)CC1', 'COc1ccc2[nH]cc(CCNC(=O)CC3CCN(Cc4ccccc4C)CC3)c2c1', 'Cc1cccc(CN2CCC(CC(=O)NCCc3c[nH]c4ccccc34)CC2)c1', 'COc1ccc2[nH]cc(CCNC(=O)CC3CCN(Cc4cccc(C)c4)CC3)c2c1', 'Cc1ccc(CN2CCC(CC(=O)NCCc3c[nH]c4ccccc34)CC2)cc1', 'COc1ccc2[nH]cc(CCNC(=O)CC3CCN(Cc4ccc(C)cc4)CC3)c2c1', 'O=C(NCCc1c[nH]c2ccccc12)C1CCN(Cc2ccccc2)CC1', 'COc1ccc2[nH]cc(CCNC(=O)C3CCN(Cc4ccccc4)CC3)c2c1', 'Cc1ccccc1CN1CCC(C(=O)NCCc2c[nH]c3ccccc23)CC1', 'COc1ccc2[nH]cc(CCNC(=O)C3CCN(Cc4ccccc4C)CC3)c2c1', 'O=C(CCc1c[nH]c2ccccc12)NC1CCN(Cc2ccccc2)CC1', 'O=C(CCc1c[nH]c2ccccc12)NCCC1CCN(Cc2ccccc2)CC1', 'O=C(CCc1c[nH]c2ccccc12)NCCC1CCN(Cc2ccccc2)CC1', 'CCCCCCCC1OCc2c[n+](C)c(C)c(OC(=O)N(C)C)c2CO1.[Br-]', 'CCCCCCCCC1OCc2c[n+](C)c(C)c(OC(=O)N(C)C)c2CO1.[Br-]', 'CCCCCCCCCC(C)C1OCc2c[n+](C)c(C)c(OC(=O)N(C)C)c2CO1.[Br-]', 'NC(=O)c1cc[n+](CC[n+]2cccc(NC(=O)/C=N\\O)c2)cc1.[Br-].[Br-]', 'NC(=O)c1cc[n+](CCC[n+]2cccc(NC(=O)/C=N\\O)c2)cc1.[Br-].[Br-]', 'NC(=O)c1cc[n+](CCC[n+]2cccc(NC(=O)/C=N\\O)c2)cc1.[Br-].[Br-]', 'NC(=O)c1cc[n+](CCCC[n+]2cccc(NC(=O)/C=N\\O)c2)cc1.[Br-].[Br-]', 'NC(=O)c1cc[n+](CCCC[n+]2cccc(NC(=O)/C=N\\O)c2)cc1.[Br-].[Br-]', 'NC(=O)c1cc[n+](CCCCC[n+]2cccc(NC(=O)/C=N\\O)c2)cc1.[Br-].[Br-]', 'NC(=O)c1cc[n+](CCCCC[n+]2cccc(NC(=O)/C=N\\O)c2)cc1.[Br-].[Br-]', 'CN(CCCOc1ccc2c3c(c(=O)oc2c1)CCCC3)Cc1cccc(OC(=O)NCCCCCc2ccccc2)c1', 'Cc1cc(=O)oc2cc(OCCCN(C)Cc3cccc(OC(=O)NCCCCCc4ccccc4)c3)ccc12', 'CN(CCCOc1ccc2c(c1)oc(=O)c1ccccc12)Cc1cccc(OC(=O)NCCCCCc2ccccc2)c1', 'CN(CCCOc1ccc2cc(-c3ccccc3)c(=O)oc2c1)Cc1cccc(OC(=O)NCCCCCc2ccccc2)c1', 'COc1ccc(-c2cc3ccc(OCCCN(C)Cc4cccc(OC(=O)NCCCCCc5ccccc5)c4)cc3oc2=O)cc1', 'COc1ccc(-c2cc3cc(OCCCN(C)Cc4cccc(OC(=O)NCCCCCc5ccccc5)c4)ccc3oc2=O)cc1', 'CN(CCCOc1ccc2cc(-c3ccc(Cl)cc3)c(=O)oc2c1)Cc1cccc(OC(=O)NCCCCCc2ccccc2)c1', 'CCCCCCCNC(=O)Oc1cccc(CN(C)CCCOc2ccc3c4c(c(=O)oc3c2)CCCC4)c1', 'CCCCCCc1ccc(NC(=O)Oc2cccc(CN(C)CCCOc3ccc4ccc(=O)oc4c3)c2)cc1', 'CN(CCCOc1ccc2ccc(=O)oc2c1)Cc1cccc(OC(=O)NCCCCCCc2ccccc2)c1', 'CN(CCCCCOc1ccc2ccc(=O)oc2c1)Cc1cccc(OC(=O)NCCCCCc2ccccc2)c1', 'CCCCCCCNC(=O)Oc1cccc(CN(C)CCCCCOc2ccc3ccc(=O)oc3c2)c1', 'CCCCCCCNC(=O)Oc1cccc(CN(C)CCCCCOc2ccc3c4c(c(=O)oc3c2)CCCC4)c1', 'CCCCCCCNC(=O)Oc1cccc(CN(C)CCCCCOc2ccc3c(c2)Oc2ncccc2C3)c1', 'O=C(NCCCCCc1ccccc1)Oc1ccc(Cn2cc(COc3ccc4ccc(=O)oc4c3)nn2)cc1', 'O=C(NCCCCCc1ccccc1)Oc1cccc(Cn2cc(COc3ccc4ccc(=O)oc4c3)nn2)c1', 'CCCCCCCNC(=O)Oc1ccc(Cn2cc(COc3ccc4ccc(=O)oc4c3)nn2)cc1', 'CCCCCCCNC(=O)Oc1cccc(Cn2cc(COc3ccc4ccc(=O)oc4c3)nn2)c1', 'O=C(NCCCCCc1ccccc1)Oc1ccc(-n2cc(COc3ccc4ccc(=O)oc4c3)nn2)cc1', 'O=C(NCCCCCc1ccccc1)Oc1cccc(-n2cc(COc3ccc4ccc(=O)oc4c3)nn2)c1', 'CCCCCCCNC(=O)Oc1cccc(-n2cc(COc3ccc4ccc(=O)oc4c3)nn2)c1', 'O=S(=O)(c1ccccc1)n1ccc2c(N3CCN(CCCCCCCCNc4c5c(nc6ccccc46)CCCC5)CC3)cccc21', 'O=S(=O)(c1ccccc1)n1ccc2c(N3CCN(CCCNc4c5c(nc6ccccc46)CCCC5)CC3)cccc21', 'O=S(=O)(c1ccccc1)n1ccc2c(N3CCN(CCCCCCCNc4c5c(nc6ccccc46)CCCC5)CC3)cccc21', 'O=S(=O)(c1ccccc1)n1ccc2c(N3CCN(CCCCCNc4c5c(nc6ccccc46)CCCC5)CC3)cccc21', 'O=S(=O)(c1ccccc1)n1ccc2c(N3CCN(CCNc4c5c(nc6ccccc46)CCCC5)CC3)cccc21', 'O=S(=O)(c1ccccc1)n1ccc2c(N3CCN(CCCCCCNc4c5c(nc6ccccc46)CCCC5)CC3)cccc21', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccc4ccccc4n3)CC1)C2', 'COC(=O)c1cccc(CN2CCC(CC3Cc4cc(OC)c(OC)cc4C3=O)CC2)n1', 'COc1cccc2c1C=[N+](c1ccccc1Br)CC2.[Br-]', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3C(F)(F)F)CC1)C2', 'COc1cc(/C=C/C(=O)Nc2ccc(CN(C)Cc3ccccc3)cc2)ccc1O', 'CC1=C(C(=O)OCCN(C)C)C(c2ccccc2)c2c(nc3c(c2N)CC(F)(F)CC3)N1', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ncccc3C)CC1)C2', 'O=C(/C=C/c1cccc(O)c1)NCCC1CCN(Cc2ccccc2)CC1', 'C#CCN(CCC1CCN(Cc2ccccc2)CC1)Cc1ccc2cccc(O)c2n1', 'COc1ccc2nc(NCCc3cc[n+](Cc4ccccc4)cc3)sc2c1', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccc(C(F)(F)F)cc3)CC1)C2', 'COc1cccc2c1C=[N+](c1ccc(F)cc1)CC2.[Br-]', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccccc1OC)CC(C)CC3', 'C=CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1cccc(F)c1)CC(C)(C)CC3', 'COc1cccc2c1C=[N+](c1ccccc1C)CC2.[Br-]', 'COc1cccc2c1C=[N+](c1ccc(I)cc1F)CC2.[Br-]', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ncc(C)cc3C)CC1)C2', 'O=[N+]([O-])c1ccc2nc(NCCc3cc[n+](Cc4ccc(F)cc4)cc3)sc2c1', 'CC1=C(C(=O)OC(C)C)C(c2cncnc2)c2c(nc3c(c2N)CC(C)(C)CC3)N1', 'C#CCN(Cc1ccc2cccc(O)c2n1)C(C#N)C1CCN(Cc2ccccc2)CC1', 'Cc1ccc(/C=C2/CNCC3=C2N=C2SC=C(c4ccc(C)cc4)N2C3c2ccc(C)cc2)cc1', 'COc1cccc([N+]2=Cc3c(cccc3OC)CC2)c1.[Br-]', 'COc1ccc(/C=C/C(=O)NCCC2CCN(Cc3ccccc3)CC2)cc1', 'COc1ccc2nc(NCCc3cc[n+](Cc4c(F)cccc4F)cc3)sc2c1', 'CCOC(=O)C1=C(C(F)(F)F)Nc2nc3c(c(N)c2C1c1cccc([N+](=O)[O-])c1)CC(C)(C)CC3', 'C#CCN(Cc1ccc2cccc(O)c2n1)C(C#N)CCC1CCN(Cc2ccccc2)CC1', 'C#CCN(Cc1ccc2cccc(O)c2n1)C(C#N)CCC1CCN(Cc2ccccc2)CC1', 'C#CCN(Cc1ccc2cccc(O)c2n1)C(C#N)CCC1CCN(Cc2ccccc2)CC1', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccn3)CC1)C2', 'Cc1ccc(/C=C2/CNCC3=C2N=C2SC=C(c4ccc5ccccc5c4)N2C3c2ccc(C)cc2)cc1', 'CCOC(=O)C1=C(C)N(C)c2nc3c(c(N)c2C1c1ccccc1Cl)CC(C)(C)CC3', 'CC(=O)/C(=N\\O)C(=O)Nc1cnc2c(c1)CC(N(C)CCc1ccccc1)CC2', 'COc1cccc2c1C=[N+](c1ccc(Cl)cc1)CC2.[Br-]', 'O=C(/C=N\\O)c1ccc(OCC2CCN(Cc3ccccc3)CC2)nc1', 'COc1cccc2c1C=[N+](c1ccc(C(F)(F)F)cc1)CC2.[Br-]', 'COc1ccc(/C=C/C(=O)NCCC2CCN(Cc3ccccc3)CC2)c(OC)c1', 'COc1ccc(/C=C/C(=O)Nc2ccc(CN(C)Cc3ccccc3)cc2)cc1OC', 'COc1ccc(/C=C/C(=O)Nc2ccc(CN(C)Cc3ccccc3)cc2)cc1OC', 'COc1cccc(/C=C/C(=O)Nc2ccc(CN(C)Cc3ccccc3)cc2)c1', 'O=C1NCCCc2c1c1ccccc1n2CCc1ccc(F)cc1', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3cccc(C)n3)CC1)C2', 'CC1=C(C(=O)OC(C)C)C(c2ccccc2Cl)c2c(nc3c(c2N)CC(C)(C)CC3)N1', 'COc1cccc2c1C=[N+](c1cccc(C#N)c1)CC2.[Br-]', 'C#CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1cccc(F)c1)CC(C)(C)CC3', 'CN(C)C(=O)Oc1cccc(C(=O)/C=C/c2ccc(O)cc2)c1', 'CN(C)C(=O)Oc1cccc(C(=O)/C=C/c2ccc(O)cc2)c1', 'Cc1ccc(C2=CSC3=NC4=C(CNC/C4=C/c4ccc(Br)cc4)C(c4ccc(Br)cc4)N23)cc1', 'O=C1c2ccccc2S(=O)(=O)N1CCCCCNCc1ccccc1', 'COC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1cccc([N+](=O)[O-])c1)CCC(C)C3', 'COc1cccc2c1C=[N+](c1ccc(Cl)cc1Cl)CC2.[Br-]', 'COc1ccc([N+]2=Cc3c(cccc3OC)CC2)cc1.[Br-]', 'COc1cccc2c1C=[N+](c1ccccc1I)CC2.[Br-]', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccc(Cl)cc3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3cccc(F)c3)CC1)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3cccc(F)n3)CC1)C2', 'COC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1cccc([N+](=O)[O-])c1)CSC3', 'Cc1ccc(/C=C2\\CN(Cc3ccc4ccccc4c3)C/C(=C\\c3ccc(C)cc3)C2=O)cc1', 'Clc1ccc(/C=C2/CNCC3=C2N=C2SC=C(c4ccc5ccccc5c4)N2C3c2ccc(Cl)cc2)cc1', 'O=C(COc1ccccc1)N1CCN(C2CCN(Cc3ccccc3)CC2)CC1', 'C#CCN(Cc1ccc2cccc(O)c2n1)CC1CCN(Cc2ccccc2)CC1', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3Cl)CC1)C2', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1cccc(Cl)c1)CC(C)(C)CC3', 'COc1cccc2c1C=[N+](c1ccc(I)cc1)CC2.[Br-]', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccc4cccc(O)c4n3)CC1)C2', 'COc1ccc(/C=C/C(=O)Nc2ccc(CN(C)Cc3ccccc3)cc2)c(OC)c1', 'COc1ccc2nc(NCCc3cc[n+](Cc4ccc(F)cc4)cc3)sc2c1', 'COC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccccc1)CSCC3', 'CC1=C(C(=O)OC(C)C)C(c2cccc(Cl)c2)c2c(nc3c(c2N)CC(C)(C)CC3)N1', 'CC1=C(C(=O)OC(C)C)C(c2ccccc2F)c2c(nc3c(c2N)CC(C)(C)CC3)N1', 'COC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1cccc([N+](=O)[O-])c1)CCCC3Cl', 'COc1cccc2c1C=[N+](c1ccccc1F)CC2.[Br-]', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccccc3F)CC1)C2', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1cccc([N+](=O)[O-])c1)CC(C)(C)CC3', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3cccc(Cl)n3)CC1)C2', 'Cc1ccc(C2=CSC3=NC4=C(CNC/C4=C/c4ccc(Cl)cc4)C(c4ccc(Cl)cc4)N23)cc1', 'COc1cc(/C=C/C(=O)Nc2cccc(CN(C)Cc3ccccc3)c2)ccc1O', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccc(C(F)(F)F)cc1)CC(C)CC3', 'COc1ccc(/C=C/C(=O)c2cccc(OC(=O)N(C)C)c2)cc1', 'Brc1ccc(/C=C2/CNCC3=C2N=C2SC=C(c4ccc5ccccc5c4)N2C3c2ccc(Br)cc2)cc1', 'CN(Cc1ccccc1)Cc1ccc(NC(=O)/C=C/c2cccc(O)c2)cc1', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1cccc([N+](=O)[O-])c1)CC(Cl)(Cl)CC3', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccc(O)c4ncccc34)CC1)C2', 'CC(C)Cc1ccc(/C=C2\\Oc3ccccc3C2=O)cc1', 'Fc1ccc(C[n+]2ccc(CCNc3nc4ccc(F)cc4s3)cc2)cc1', 'O=C1/C(=C/c2ccccc2)CN(Cc2ccc3ccccc3c2)C/C1=C\\c1ccccc1', 'S=C1/C(=C/c2ccccc2)Oc2ccccc21', 'COc1cccc2c1C=[N+](c1ccc(C)cc1)CC2.[Br-]', 'Brc1ccc(/C=C2/CNCC3=C2N=C2SC=C(c4ccccc4)N2C3c2ccc(Br)cc2)cc1', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1cccc(C(F)(F)F)c1)CC(C)(C)CC3', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3cccc(Cl)c3)CC1)C2', 'C=CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1cccc([N+](=O)[O-])c1)CC(C)(C)CC3', 'C#CCN(CCCC1CCN(Cc2ccccc2)CC1)Cc1ccc2cccc(O)c2n1', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3cccc(C(F)(F)F)c3)CC1)C2', 'CCN1CCN(C2CCN(C(=O)COc3ccc(-c4ccccc4)cc3)CC2)CC1', 'CCN1CCN(C2CCN(C(=O)COc3ccc(-c4ccccc4)cc3)CC2)CC1', 'CC1=C(C(=O)OC2CCC2)C(c2ccccc2F)c2c(nc3c(c2N)CC(C)(C)CC3)N1', 'O=C(/C=C/c1ccc(O)cc1O)NCCC1CCN(Cc2ccccc2)CC1', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1cccc(F)c1)CC(C)(C)CC3', 'COc1cccc2c1C=[N+](c1ccccc1)CC2.[Br-]', 'O=C1NCCCc2c1c1ccccc1n2CCc1ccccc1', 'C#CCN(Cc1ccc2cccc(O)c2n1)C(C#N)CC1CCN(Cc2ccccc2)CC1', 'CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1cccc([N+](=O)[O-])c1)CC(F)(F)CC3', 'Cc1ccc(/C=C2/CNCC3=C2N=C2SC=C(c4ccc(Br)cc4)N2C3c2ccc(C)cc2)cc1', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3cccc(Br)n3)CC1)C2', 'COc1cccc2c1C=[N+](c1cccc(C)c1)CC2.[Br-]', 'Cc1cccc(CN2C/C(=C\\c3ccccc3)C(=O)/C(=C/c3ccccc3)C2)c1', 'COc1ccc(/C=C/C(=O)Nc2ccc(CN(C)Cc3ccccc3)cc2)cc1', 'COc1ccc(/C=C/C(=O)Nc2ccc(CN(C)Cc3ccccc3)cc2)cc1', 'COc1cccc2c1C=[N+](c1ccc(Br)cc1F)CC2.[Br-]', 'COc1ccc(CN2CCCCC2COc2ccc(C(=O)/C=N\\O)cn2)cc1OC', 'COc1cccc2c1C=[N+](c1c(F)cccc1F)CC2.[Br-]', 'C#CCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1cccc([N+](=O)[O-])c1)CC(C)(C)CC3', 'COCCOC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccccc1Cl)CC(C)(C)CC3', 'COc1cccc2c1C=[N+](c1ccc(Br)cc1Br)CC2.[Br-]', 'O=C1/C(=C/c2ccc(Br)cc2)CN(Cc2ccc3ccccc3c2)C/C1=C\\c1ccc(Br)cc1', 'CC1=C(C(=O)OCC2CC2)C(c2ccccc2Cl)c2c(nc3c(c2N)CC(C)(C)CC3)N1', 'Cc1ccc(/C=C2\\CN(Cc3cccc(C)c3)C/C(=C\\c3ccc(C)cc3)C2=O)cc1', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3ccc(F)cc3)CC1)C2', 'CC1=C(C(=O)OC(C)C)C(c2cccc(F)c2)c2c(nc3c(c2N)CC(C)(C)CC3)N1', 'COC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1ccccc1)CC(C)(C)CC3', 'COc1cccc2c1C=[N+](c1ccc(Br)cc1)CC2.[Br-]', 'COC(=O)C1=C(C)Nc2nc3c(c(N)c2C1c1cccc([N+](=O)[O-])c1)CC(C)(C)CC3', 'COc1ccc(CN(CC(=O)NCc2ccccc2)C(=O)CCCCCn2cc[n+](C)c2/C=N\\O)cc1.[I-]', 'COc1ccc(CN(CC(=O)NCc2ccccc2)C(=O)CCCCCn2cc[n+](C)c2/C=N\\O)cc1.[I-]', 'COc1ccc(CN(CC(=O)NCc2ccccc2)C(=O)CCCCCn2cc[n+](C)c2/C=N\\O)cc1.[I-]', 'COc1ccc(CN(CC(=O)NCc2ccccc2)C(=O)CCCCCn2cc[n+](C)c2/C=N\\O)cc1.[I-]', 'CCN(CC)CCCCCCCCOc1ccc(/C=C/C(=O)O)cc1OC', 'CCOC(=O)[C@@H](CCC(=O)NCCC1CCN(Cc2ccccc2)CC1)NC(=O)c1ccccc1', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(CN3CCOCC3)c(O)c1)O2', 'COc1cc2c(cc1OC)C(=O)C(Cc1cc[n+](Cc3ccccc3)cc1C(C)=O)C2.[Br-]', 'COc1ccc(CCNc2nc(NC(C)=O)nc3ccccc23)cc1OC', 'O=C(CCCC1CCSS1)NCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'COc1ccc(CNc2nc(NC(N)=O)nc3ccccc23)cc1OC', 'N#CC1=C(N)N(Cc2ccccc2)C(NCc2ccccc2)=C([N+](=O)[O-])C12c1ccccc1-c1nc3ccc([N+](=O)[O-])cc3nc12', 'CN1CC2C(=O)N(Cc3ccccc3)C(=O)C2C12C(=O)c1ccccc1C2=O', 'CCN(CC)c1ccc(/C=C/c2ccc3ccccc3[n+]2C)c(O)c1.[I-]', 'COc1cc2c(cc1OC)C(=O)C(CC1C=CN(Cc3ccccc3C)C=C1C(C)=O)C2', 'CC(=O)Nc1nc(NCc2ccccc2)c2ccccc2n1', 'CCNc1nc(NCCc2ccccc2)c2ccccc2n1', 'Cc1cc2nc3c(nc2cc1C)C1(C(C#N)=C(N)N2CCNC2=C1[N+](=O)[O-])c1ccccc1-3', 'CCOC(=O)/C=C/c1ccc(OCCCCCCN2CCCC2)c(OC)c1', 'COc1ccc(NCc2ccc3c(c2)C(=O)C(=O)N3C(=O)c2ccccc2)cc1', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCCCCCN(C)C', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1cc[n+](Cc3ccccc3Br)cc1)C2.[Br-]', 'C[n+]1c2ccccc2c(N)c2ccccc21.F[B-](F)(F)F', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCCCN1CCCC1', 'CN1CCN(Cc2ccc3c(c2)C(=O)C(=O)N3C(=O)c2ccccc2)CC1', 'COc1cc2c(cc1OC)CN(CC1CCN(Cc3cc(F)cc(/C=N/O)c3O)CC1)CC2', 'O=C(CC[C@@H](NC(=O)c1ccccc1)C(=O)O)NCCC1CCN(Cc2ccccc2)CC1', 'CNc1nc(NCCc2ccccc2)c2ccccc2n1', 'CCN(CC)CCOc1ccc(/C=C/C(=O)Nc2ccccc2)cc1OC', 'CC1(C)CNC2=C([N+](=O)[O-])C3(C(C#N)=C(N)N2C1)c1ccccc1-c1nc2ccccc2nc13', 'CN1c2ccccc2C(N2CCSCC2)c2ccccc21', 'CN1c2ccccc2C(N2CCSCC2)c2ccccc21', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCCN(C)C', 'Oc1cc(O)cc(CCc2ccc(Nc3c4c(nc5cc(Cl)ccc35)CCCC4)cc2)c1', 'CCO[P@](=O)(SCC)c1ccccc1', 'CCO[P@](=O)(SCC)c1ccccc1', 'CCN(CC)c1ccc(/C=C/c2ccc3ccccc3[n+]2C)cc1.[I-]', 'COc1cc2c(cc1OC)C(=O)C(Cc1cc[n+](Cc3ccccc3Cl)cc1C(C)=O)C2.[Br-]', 'COc1ccccc1N1C(=O)C2CN(C)C3(C(=O)c4ccccc4C3=O)C2C1=O', 'CCN(CCCN1C(=O)c2ccccc2C1=O)Cc1ccccc1OC', 'CCOC(=O)CC[C@H](NC(=O)OCc1ccccc1)C(=O)NCCC1CCN(Cc2ccccc2)CC1', 'c1ccc(CNc2nc(NC3CC3)nc3ccccc23)cc1', 'Nc1c2c(nc3c1C(c1ccccc1)c1ccc4cccnc4c1O3)CCCC2', 'CCN(CC)CCCCCOc1ccc(/C=C/C(=O)Nc2ccccc2)cc1OC', 'C[n+]1c(/C=C/c2cc3ccccc3[nH]2)cc(N2CCOCC2)c2ccccc21.[I-]', 'COC(=O)c1c[n+](Cc2ccccc2)ccc1CC1Cc2cc(OC)c(OC)cc2C1=O.[Br-]', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCCN1CCCCC1', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCC(C)[C@@H]2C3C)cc1', 'COc1ccc(CN(CC(=O)NC(C)C)C(=O)CCCCCn2cc[n+](C)c2/C=N\\O)cc1.[I-]', 'COc1ccc(CN(CC(=O)NC(C)C)C(=O)CCCCCn2cc[n+](C)c2/C=N\\O)cc1.[I-]', 'COc1ccc(CN(CC(=O)NC(C)C)C(=O)CCCCCn2cc[n+](C)c2/C=N\\O)cc1.[I-]', 'CCN(CC)Cc1ccc(NC(=O)c2ccccc2)c(C(=O)C(=O)N(CC)CC)c1', 'CCN(CCCCCCNC(=O)COc1ccc(-c2cc(=O)c3c(OC)c(OC)c(OC)cc3o2)cc1)Cc1ccccc1OC', 'CCc1ccc(N2C(=O)C3CN(C)C4(C(=O)c5ccccc5C4=O)C3C2=O)cc1', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(O)c(CN(C)C)c1)O2', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCCCN1CCCCC1', 'NC(=O)Cc1nc(NCCc2ccccc2)c2ccccc2n1', 'Cc1ccc(-c2c3ccccc3[n+](C)c3ccccc23)cc1.F[B-](F)(F)F', 'CCN(CCCCN1C(=O)c2ccccc2C1=O)Cc1ccccc1OC', 'CC(=O)Nc1ccc(-c2c3ccccc3[n+](C)c3ccccc23)cc1.F[B-](F)(F)F', 'C[n+]1c2ccccc2c(-c2cccs2)c2ccccc21.F[B-](F)(F)F', 'CN1CCN(c2ccc(/C=C/c3ccc4ccccc4[n+]3C)cc2)CC1.[I-]', 'COc1ccc(CNc2c3c(nc4cc(Cl)ccc24)CCCC3)cc1OC', 'NC(=O)Nc1nc(NCCc2ccccc2)c2ccccc2n1', 'CC(C)Nc1nc(NCCc2ccccc2)c2ccccc2n1', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1cc[n+](Cc3ccc([N+](=O)[O-])cc3)cc1)C2.[Br-]', 'N#CC1=C(N)N2CCCNC2=C([N+](=O)[O-])C12c1ccccc1-c1nc3ccccc3nc12', 'Cc1cc(-c2c3ccccc3[n+](C)c3ccccc23)cc(C)c1C.F[B-](F)(F)F', 'COc1cc2c(cc1OC)C(=O)C(Cc1cc[n+](Cc3ccccc3)cc1C#N)C2.[Br-]', 'CN1CC2C(=O)N(c3ccc(F)cc3)C(=O)C2C12C(=O)c1ccccc1C2=O', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCCCCCN1CCCC1', 'CCOC(=O)/C=C/c1ccc(OCCCCCCN(CC)CC)c(OC)c1', 'CSP(C)(=O)OC1CCCCC1', 'COc1ccc2nc3c(c(NCCCCCCCNC(=S)NC45CC6CC(CC(C6)C4)C5)c2c1)CCCC3', 'COc1ccc(N2C(=O)C3CN(C)C4(C(=O)c5ccccc5C4=O)C3C2=O)cc1', 'CCN(Cc1ccc(/C=C2\\Oc3cc(OC)c(OC)cc3C2=O)cc1O)Cc1ccccc1N(C)C', 'CC(=O)c1c[n+](Cc2ccccc2C)ccc1CC1Cc2cc(C)ccc2C1=O.[Br-]', 'NC(=O)Cc1nc(NCc2ccccc2)c2ccccc2n1', 'COc1ccc2cc(C(=O)N[C@@H](CCC(=O)NCCC3CCN(Cc4ccccc4)CC3)C(=O)O)ccc2c1', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1cc[n+](Cc3ccc(F)cc3)cc1)C2.[Br-]', 'COc1ccccc1NC(=S)NCCN1CCCCC1.Cl', 'CC1(C)CNC2=C([N+](=O)[O-])C3(C(C#N)=C(N)N2C1)c1ccccc1-c1nc2ccc([N+](=O)[O-])cc2nc13', 'CC(=O)c1c[n+](Cc2ccccc2C)ccc1CC1Cc2cc(C)c(C)cc2C1=O.[Br-]', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCCCCCCCCCN1CCCCC1', 'N#CC1=C(N)N(Cc2ccccc2)C(NCc2ccccc2)=C([N+](=O)[O-])C12c1ccccc1-c1nc3ccccc3nc12', 'Cc1ccc2c(c1)C(=O)C(=O)N2Cc1ccccc1[N+](=O)[O-]', 'COc1cc(/C=C/C(=O)O)ccc1OCCCCCCN1CCCCC1', 'Clc1ccc2c(NCc3ccccn3)c3c(nc2c1)CCCC3', 'COc1ccc(NC(=S)NCCN(C)C)cc1Cl', 'CCN(CC)Cc1ccc(/C=C2\\Oc3cc(OC)c(OC)cc3C2=O)cc1O', 'CN1c2ccccc2C(n2cnc([N+](=O)[O-])n2)c2ccccc21', 'C#CCN(CCCN1C(=O)c2ccccc2C1=O)Cc1ccccc1', 'CN1c2ccccc2C(n2cncn2)c2ccccc21', 'CN1c2ccccc2C(n2cncn2)c2ccccc21', 'COc1ccc(CNc2c3c(nc4ccccc24)CCCC3)cc1OC', 'CCCCCCOC(=O)[C@H](CCC(=O)NCCC1CCN(Cc2ccccc2)CC1)NC(=O)c1c[nH]c2ccccc12', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1cc[n+](Cc3ccccc3F)cc1)C2.[Br-]', 'COc1cc2c(cc1OC)C(=O)C(Cc1cc[n+](Cc3ccccc3C)cc1C(N)=O)C2.[Br-]', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1cc[n+](Cc3ccc(C)cc3)cc1)C2.[Br-]', 'CCNc1nc(NCc2ccc(OC)c(OC)c2)c2ccccc2n1', 'O=C(Nc1ccc(CN2CCOCC2)cc1C(=O)C(=O)N1C(=O)CCC1=O)c1ccccc1', 'CN1c2ccccc2C(N2CCOCC2)c2ccccc21', 'CN1c2ccccc2C(N2CCOCC2)c2ccccc21', 'COc1cc2c(cc1OC)CN(CC1CCN(Cc3cc(Br)cc(/C=N/O)c3O)CC1)CC2', 'COC(=O)c1cc2c(OC(=O)N(C)C)cccc2[n+](C)c1', 'C[n+]1c2ccccc2c(-c2ccc(F)cc2)c2ccccc21.F[B-](F)(F)F', 'O=C(CC[C@H](NC(=O)c1ccccc1)C(=O)O)NCCC1CCN(Cc2ccccc2)CC1', 'CCNc1nc(NCCc2ccc(OC)c(OC)c2)c2ccccc2n1', 'CCN(CC)CCCOc1ccc(/C=C/C(=O)Nc2ccccc2)cc1OC', 'NC(=O)c1cc[n+](COC[n+]2ccc(/C=N/O)cc2)cc1.[Cl-].[Cl-]', 'NC(=O)c1cc[n+](COC[n+]2ccc(/C=N/O)cc2)cc1.[Cl-].[Cl-]', 'NC(=O)c1cc[n+](COC[n+]2ccc(/C=N/O)cc2)cc1.[Cl-].[Cl-]', 'CCNc1nc(NCc2ccccc2)c2ccccc2n1', 'CCCCCCOC(=O)[C@@H](CCC(=O)NCCC1CCN(Cc2ccccc2)CC1)NC(=O)c1ccccc1', 'CN1CC2C(=O)N(c3ccccc3)C(=O)C2C12C(=O)c1ccccc1C2=O', 'NS(=O)(=O)c1ccc(NC(=S)NCCN2CCCCC2)cc1', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCCCCCN1CCCCC1', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCCCCCN1CCCCC1', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCCCCCN1CCCCC1', 'CN1CC2C(=O)N(c3cccc(F)c3)C(=O)C2C12C(=O)c1ccccc1C2=O', 'CCN(CC)c1ccc(/C=C/c2ccc3ccccc3[n+]2C)c(Cl)c1.[I-]', 'CCOC(=O)[C@H](CCC(=O)NCCC1CCN(Cc2ccccc2)CC1)NC(=O)c1ccc2cc(OC)ccc2c1', 'CCN(CC)c1ccc(-c2c3ccccc3[n+](C)c3ccccc23)cc1.F[B-](F)(F)F', 'CCN(CC)c1ccc(-c2c3ccccc3[n+](C)c3ccccc23)cc1.F[B-](F)(F)F', 'NC(=O)Nc1nc(NCc2ccccc2)c2ccccc2n1', 'CCCCN(CC(=O)NCc1ccccc1)C(=O)CCCCCn1cc[n+](C)c1/C=N\\O.[I-]', 'CCCCN(CC(=O)NCc1ccccc1)C(=O)CCCCCn1cc[n+](C)c1/C=N\\O.[I-]', 'CCOC(=O)[C@H](CCC(=O)NCCC1CCN(Cc2ccccc2)CC1)NC(=O)c1c[nH]c2ccccc12', 'Oc1cc(O)cc(/C=C/c2ccc(Nc3c4c(nc5cc(Cl)ccc35)CCCC4)cc2)c1', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(CN3CCN(C)CC3)c(O)c1)O2', 'COc1cc2c(cc1OC)C(=O)C(Cc1cc[n+](Cc3ccccc3C)cc1C(C)=O)C2.[Br-]', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1cc[n+](Cc3cccc(Br)c3)cc1)C2.[Br-]', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCCCCCCCCCN1CCCC1', 'CCOc1ccc(Cc2nc3cc(C(=O)N(CC)CC)ccc3n2CCCCCCN2CCCCC2)cc1', 'CCOc1ccc(Cc2nc3cc(C(=O)N(CC)CC)ccc3n2CCCCCCN2CCCCC2)cc1', 'O=C1C(=O)N(Cc2ccccc2[N+](=O)[O-])c2ccc(Br)cc21', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1cc[n+](Cc3ccc(Br)cc3)cc1)C2.[Br-]', 'COc1cc(CCc2ccc(Nc3c4c(nc5cc(Cl)ccc35)CCCC4)cc2)cc(OC)c1', 'CCCNc1nc(NCc2ccc(OC)c(OC)c2)c2ccccc2n1', 'CCCNc1nc(NCCc2ccccc2)c2ccccc2n1', 'N#CC1=C(N)N2CCNC2=C([N+](=O)[O-])C12c1ccccc1-c1nc3ccc([N+](=O)[O-])cc3nc12', 'CC(=O)c1c[n+](Cc2ccccc2C)ccc1CC1Cc2cc3c(cc2C1=O)OCO3.[Br-]', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCCCCCCCN(C)C', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1cc[n+](Cc3cccc(C)c3)cc1)C2.[Br-]', 'CC(=O)Nc1nc(NCCc2ccccc2)c2ccccc2n1', 'COc1cc(/C=C/C(=O)N2CCC(Cc3ccccc3)CC2)ccc1OCCCCN1CCc2ccccc2C1', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCCN1CCCC1', 'O=C1C(=O)N(C(=O)c2ccccc2)c2ccc(CN3CCCCC3)cc21', 'CCOc1ccc(Cc2nc3cc(C(=O)N(CC)CC)ccc3n2CCCCN2CCCCC2)cc1', 'CCOc1ccc(Cc2nc3cc(C(=O)N(CC)CC)ccc3n2CCCCN2CCCCC2)cc1', 'COc1cc(/C=N/O)c(O)c(CN2CCC(CN3CCc4cc(OC)c(OC)cc4C3)CC2)c1', 'COc1ccc(C2c3ccc4cccnc4c3Oc3nc4c(c(N)c32)CCCC4)cc1', 'O=C(CC[C@H](NC(=O)c1c[nH]c2ccccc12)C(=O)O)NCCC1CCN(Cc2ccccc2)CC1', 'COc1ccc(CCNc2nc(NC3CC3)nc3ccccc23)cc1OC', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCCCCCCCCCN(C)C', 'COc1cc(/C=C/c2cc(N3CCN(C)CC3)c3ccccc3[n+]2C)ccc1N1CCN(C)CC1.[I-]', 'CCN(CC)CCCCCCCCCCOc1ccc(/C=C/C(=O)Nc2ccccc2)cc1OC', 'CCOC(=O)/C=C/c1ccc(OCCCCCCCCN(CC)CC)c(OC)c1', 'CN1C(=O)C2CN(C)C3(C(=O)c4ccccc4C3=O)C2C1=O', 'Cc1cc(-c2c3ccccc3[n+](C)c3ccccc23)cc(C)c1O.F[B-](F)(F)F', 'CN1c2ccccc2C(n2nnc3ccccc32)c2ccccc21', 'CN1c2ccccc2C(n2nnc3ccccc32)c2ccccc21', 'CN1CC2C(=O)N(C3CCCCC3)C(=O)C2C12C(=O)c1ccccc1C2=O', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN[C@@H]2N3)cc1', 'CC(C)c1ccc(NC(=O)Oc2ccc3c(c2)[C@]2(C)CCN[C@@H]2N3)cc1', 'CCN(CCCCCCN1C(=O)c2ccccc2C1=O)Cc1ccccc1OC', 'COc1cc2c(cc1OC)CN(CC1CCN(Cc3cc(C)cc(/C=N/O)c3O)CC1)CC2', 'COc1cc2c(cc1OC)CN(CC1CCN(Cc3cc(C)cc(/C=N/O)c3O)CC1)CC2', 'COc1cc2c(cc1OC)CN(CC1CCN(Cc3cc(C)cc(/C=N/O)c3O)CC1)CC2', 'Cc1ccc(N2C(=O)C3CN(C)C4(C(=O)c5ccccc5C4=O)C3C2=O)cc1', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCCCCN1CCCCC1', 'C[n+]1c2ccccc2c(-c2ccccc2)c2ccccc21.F[B-](F)(F)F', 'COc1cc(/C=C/C(=O)O)ccc1OCCCCCCN1CCCC1', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(CN(C)C)c(O)c1)O2', 'O=C1C(=O)N(C(=O)c2ccccc2)c2ccc(CNc3ccc(Br)cc3)cc21', 'Clc1ccc(CNC2CCN(Cc3ccccc3)CC2)c(Cl)c1', 'COc1ccc(CNc2nc(NC(C)=O)nc3ccccc23)cc1OC', 'c1ccc(CCNc2nc(NC3CC3)nc3ccccc23)cc1', 'Cc1cc2nc3c(nc2cc1C)C1(C(C#N)=C(N)N2CCCNC2=C1[N+](=O)[O-])c1ccccc1-3', 'CCN(CCCCN1C(=O)c2ccccc2C1=O)Cc1ccccc1', 'COc1ccc(CCNc2c3c(nc4cc(Cl)ccc24)CCCC3)cc1OC', 'COc1cc2c(cc1OC)C(=O)C(Cc1cc[n+](Cc3cccc(Cl)c3)cc1C(N)=O)C2.[Br-]', 'COc1ccc(NC(=S)NCCN2CCOCC2)cc1Cl', 'CCN(CCCCCCNC(=O)COc1ccc(-c2cc(=O)c3c(O)c(OC)c(OC)cc3o2)cc1)Cc1ccccc1OC', 'N#CC1=C(N)N2CCCNC2=C([N+](=O)[O-])C12c1ccccc1-c1nc3ccc([N+](=O)[O-])cc3nc12', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCCCCN1CCCC1', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCN(C)C', 'COc1cc(CCc2ccc(Nc3c4c(nc5ccccc35)CCCC4)cc2)cc(OC)c1', 'CN1CCN(c2ccc(/C=C/c3cc(N4CCN(C)CC4)c4ccccc4[n+]3C)cc2)CC1.[I-]', 'CC(C)c1ccc(N2C(=O)C3CN(C)C4(C(=O)c5ccccc5C4=O)C3C2=O)cc1', 'COc1cc(/C=C/c2ccc3ccccc3[n+]2C)ccc1N1CCN(C)CC1.[I-]', 'CCOC(=O)[C@H](CCC(=O)NCCC1CCN(Cc2ccccc2)CC1)NC(=O)c1ccccc1', 'COc1cc2c(cc1OC)C(=O)C(Cc1cc[n+](Cc3cccc(C)c3)cc1C(C)=O)C2.[Br-]', 'COc1ccc2c(c1)CC(Cc1cc[n+](Cc3ccccc3C)cc1C(C)=O)C2=O.[Br-]', 'CCOC(=O)/C=C/c1ccc(OCCCCCCN2CCCCC2)c(OC)c1', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(CN3CCCC3)c(O)c1)O2', 'CNc1nc(NCCc2ccc(OC)c(OC)c2)c2ccccc2n1', 'CCN(CC)CCCCCCOc1ccc(/C=C/C(=O)Nc2ccccc2)cc1OC', 'Cc1cc2nc3c(nc2cc1C)C1(C(C#N)=C(N)N(Cc2ccccc2)C(NCc2ccccc2)=C1[N+](=O)[O-])c1ccccc1-3', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1cc[n+](Cc3cccc(F)c3)cc1)C2.[Br-]', 'CCn1c2ccccc2c2cc(/C=C/c3cc(N4CCOCC4)c4ccccc4[n+]3C)ccc21.[I-]', 'COc1ccc(CCNc2nc(NC(C)C)nc3ccccc23)cc1OC', 'COC(=O)c1c[n+](Cc2ccccc2C)ccc1CC1Cc2cc(OC)c(OC)cc2C1=O.[Br-]', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCN1CCCCC1', 'CC1CC[C@@]2(C)c3cc(OC(=O)Nc4ccccc4)ccc3C(C)[C@@H]12', 'COc1cccc(C2c3ccc4cccnc4c3Oc3nc4c(c(N)c32)CCCC4)c1', 'COc1cccc(C2c3ccc4cccnc4c3Oc3nc4c(c(N)c32)CCCC4)c1', 'COc1ccc(-c2c3ccccc3[n+](C)c3ccccc23)cc1.F[B-](F)(F)F', 'COc1ccc(CCNc2nc(CC(N)=O)nc3ccccc23)cc1OC', 'CCN(CC)c1ccc(/C=C/c2ccc3ccccc3[n+]2C)c(OC)n1.[I-]', 'CCN(CCCCCCN1C(=O)c2ccccc2C1=O)Cc1ccccc1', 'C[n+]1c2ccccc2c(-c2ccc(NC(=O)C(F)(F)F)cc2)c2ccccc21.F[B-](F)(F)F', 'CCN(CC)c1ccc(/C=C/c2ccc3ccccc3[n+]2C)c(OC)c1.[I-]', 'COc1ccc(CNc2nc(NC3CC3)nc3ccccc23)cc1OC', 'CN1CC2C(=O)N(c3ccc(Br)cc3)C(=O)C2C12C(=O)c1ccccc1C2=O', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCCCCCCCN1CCCCC1', 'CCN1CCN(Cc2ccc(/C=C3\\Oc4cc(OC)c(OC)cc4C3=O)cc2O)CC1', 'COc1cc2c(cc1OC)C(=O)C(Cc1cc[n+](Cc3ccccc3)cc1)C2.[Br-]', 'CNc1nc(NCc2ccc(OC)c(OC)c2)c2ccccc2n1', 'CCN(CCCCNC(=O)COc1ccc(-c2cc(=O)c3c(OC)c(OC)c(OC)cc3o2)cc1)Cc1ccccc1OC', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1cc[n+](Cc3ccccc3)cc1)C2.[Br-]', 'CN1CC2C(=O)N(c3ccc(Cl)cc3)C(=O)C2C12C(=O)c1ccccc1C2=O', 'COc1ccc(CCNc2nc(NC(N)=O)nc3ccccc23)cc1OC', 'CCN(CC)CCCCCCOc1ccc(/C=C/C(=O)O)cc1OC', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1cc[n+](Cc3ccccc3C)cc1)C2.[Br-]', 'N#CC1=C(N)N2CCNC2=C([N+](=O)[O-])C12c1ccccc1-c1nc3ccccc3nc12', 'COc1ccc(CNc2nc(NC(C)C)nc3ccccc23)cc1OC', 'CC(=O)c1c[n+](Cc2ccccc2C)ccc1CC1Cc2cc(Cl)ccc2C1=O.[Br-]', 'COc1ccc(CNc2nc(CC(N)=O)nc3ccccc23)cc1OC', 'CCN(CCCCNC(=O)COc1ccc(-c2cc(=O)c3c(O)c(OC)c(OC)cc3o2)cc1)Cc1ccccc1OC', 'Cc1cc2nc3c(nc2cc1C)C1(C(C#N)=C(N)N2CC(C)(C)CNC2=C1[N+](=O)[O-])c1ccccc1-3', 'Cc1ccc2c(c1)C(=O)C(=O)N2C(=O)N(C)C', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCCCCCCCN1CCCC1', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(CN3CCCCC3)c(O)c1)O2', 'O=C1C(=O)N(Cc2ccc(Br)cc2)c2ccc(Br)cc21', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1cc[n+](Cc3cccc([N+](=O)[O-])c3)cc1)C2.[Br-]', 'COc1cc2c(cc1OC)C(=O)C(CC1C=CN(Cc3cccc(Cl)c3)C=C1C(N)=O)C2', 'COc1cc2c(cc1OC)C(=O)C(CC1C=CN(Cc3cccc(Cl)c3)C=C1C(N)=O)C2', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCN1CCCC1', 'c1ccc(CNc2c3c(nc4ccccc24)CCCC3)nc1', 'COc1cc2c(cc1OC)CN(CC1CCN(Cc3cc(Cl)cc(/C=N/O)c3O)CC1)CC2', 'COc1cc2c(cc1OC)C(=O)C(Cc1cc[n+](Cc3cccc(Cl)c3)cc1C(C)=O)C2.[Br-]', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(CN(C)Cc3ccccc3)c(O)c1)O2', 'CCOC(=O)CC[C@@H](NC(=O)OCc1ccccc1)C(=O)NCCC1CCN(Cc2ccccc2)CC1', 'CCCNc1nc(NCCc2ccc(OC)c(OC)c2)c2ccccc2n1', 'CCN(CC)CCCCOc1ccc(/C=C/C(=O)Nc2ccccc2)cc1OC', 'COc1cc(/C=C/C(=O)Nc2ccccc2)ccc1OCCCCCN(C)C', 'Cc1ccc(C2c3ccc4cccnc4c3Oc3nc4c(c(N)c32)CCCC4)cc1', 'CN1CC2C(=O)N(c3ccc([N+](=O)[O-])cc3)C(=O)C2C12C(=O)c1ccccc1C2=O', 'Nc1c2c(nc3c1C(c1cccc(F)c1)c1ccc4cccnc4c1O3)CCCC2', 'C[n+]1c(/C=C/c2ccc(Cl)cc2)ccc2ccccc21.[I-]', 'C[n+]1c2ccccc2c(-c2ccc(N)cc2)c2ccccc21.F[B-](F)(F)F', 'C[n+]1c2ccccc2c(-c2ccc(N)cc2)c2ccccc21.F[B-](F)(F)F', 'CC(=O)C1=CN(Cc2ccccc2C)C=CC1CC1Cc2cc(C)c(C)cc2C1=O', 'CCCNc1nc(NCc2ccccc2)c2ccccc2n1', 'CCN(CC)CCCCCCCCOc1ccc(/C=C/C(=O)Nc2ccccc2)cc1OC', 'CC1=C[C@H]2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCNC(=O)c4cc(O)c5c(c4)C(=O)c4cccc(O)c4C5=O)c3[C@@H](C1)C2', 'CC1=C[C@H]2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2', 'CC1=C[C@@H]2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCNC(=O)c4cc(O)c5c(c4)C(=O)c4cccc(O)c4C5=O)c3[C@H](C1)C2', 'C=C(C)[C@H](O)CC[C@]1(C)[C@@H](CC=C(C)C)C[C@@]2(CC=C(C)C)C(=O)[C@]1(C(=O)C(C)C)C(=O)C1=C2O[C@H](C(C)(C)O)C1', 'COC(C)(C)/C=C/C[C@]1(C)[C@@H](CC=C(C)C)C[C@]2(CC=C(C)C)C(=O)C3=C(O[C@@H](C(C)(C)O)C3)[C@@]1(C(=O)C(C)C)C2=O', 'CC(C)=CCC[C@]1(C)[C@@H](CC=C(C)C)C[C@@]2(CC=C(C)C)C(=O)[C@]1(C(=O)C(C)C)C(=O)C1=C2O[C@H](C(C)(C)O)[C@H]1O', 'C=C(C)[C@@H](O)CC[C@]1(C)[C@@H](CC=C(C)C)C[C@]23C[C@@H](C(C)(C)O)OC2=C(CC=C(C)C)C(=O)[C@@]1(C(=O)C(C)C)C3=O', 'CC(C)=CC[C@H]1C[C@@]2(CC=C(C)C)C(=O)[C@](C(=O)C(C)C)(C(=O)C3=C2O[C@H](C(C)(C)O)C3)[C@]1(C)CCC(=O)C(C)C', 'C=C(C)[C@H](O)CC[C@]1(C)[C@@H](CC=C(C)C)C[C@]2(CC=C(C)C)C(=O)C3=C(O[C@H](C(C)(C)O)C3)[C@@]1(C(=O)C(C)C)C2=O', 'O=C(Cc1cc(=O)oc2cc(O)ccc12)NCCCCCCCNC(=O)c1cccc2c(=O)c3ccccc3[nH]c12', 'C=C(C)[C@@H](O)CC[C@]1(C)[C@@H](CC=C(C)C)C[C@]2(CC=C(C)C)C(=O)C3=C(O[C@H](C(C)(C)O)C3)[C@@]1(C(=O)C(C)C)C2=O', 'CC(=O)[C@@]12C(=O)C(CC=C(C)C)=C3O[C@H](C(C)(C)O)C[C@]3(C[C@H](CC=C(C)C)[C@@]1(C)CCC=C(C)C)C2=O', 'C=C(C)[C@H](O)CC[C@]1(C)[C@@H](CC=C(C)C)C[C@@]2(CC=C(C)C)C(=O)[C@]1(C(=O)C(C)C)C(=O)C1=C2O[C@@H](C(C)(C)O)C1', 'CC(C)=CCC[C@]1(C)[C@@H](CC=C(C)C)C[C@]2(CC=C(C)C)C(=O)C3=C(O[C@@H](C(C)(C)O)[C@@H]3O)[C@@]1(C(=O)C(C)C)C2=O', 'CC(C)=CCC1=C2O[C@H](C(C)(C)O)C[C@@]23C[C@H](CC=C(C)C)[C@@](C)(CCC(=O)C(C)C)[C@@](C(=O)C(C)C)(C1=O)C3=O', 'CC(C)=CCC[C@]1(C)[C@@H](CC=C(C)C)C[C@]2(CC=C(C)C)C(=O)C3=C(O[C@H](C(C)(C)O)[C@H]3O)[C@@]1(C(=O)C(C)C)C2=O', 'CC(C)=CCC[C@]1(C)[C@@H](CC=C(C)C)C[C@@]2(CC=C(C)C)C(=O)[C@]1(C(=O)C(C)C)C(=O)C1=C2O[C@@H](C(C)(C)O)[C@@H]1O', 'COC(C)(C)/C=C/C[C@]1(C)[C@@H](CC=C(C)C)C[C@]2(CC=C(C)C)C(=O)C3=C(O[C@H](C(C)(C)O)C3)[C@@]1(C(=O)C(C)C)C2=O', 'Oc1ccc(-c2oc3cc(O)c4c5c3c2-c2cc(O)cc(O)c2[C@H](c2ccc(O)cc2)[C@@H]5[C@H](c2ccc(O)cc2)[C@@H]4c2cc(O)cc(O)c2)cc1', 'COC(=O)c1ccccc1C[n+]1ccc(CC2Cc3cc(OC)c(OC)cc3C2=O)c(C(=O)Oc2c(Cl)cc(Cl)cc2Cl)c1.[Br-]', 'Cc1cc(=O)oc2cc(OCCCSC(=S)N3CCCCC3)ccc12', 'Cc1cc(=O)oc2cc(OCCSC(=S)N3CCC(N4CCCCC4)CC3)ccc12', 'COc1cc2c(cc1O)C(=O)C(Cc1cc[n+](Cc3cccc(Cl)c3)cc1C(C)=O)C2', 'Nc1c2c(nc3c1C(c1ccc(F)cc1)NC(=S)N3)CCCC2', 'COc1ccc(C(=O)CCc2cc[n+](Cc3cccc(Cl)c3)cc2C#N)cc1OC.[Br-]', 'COc1ccc2c(c1)CCC(Cc1cc[n+](Cc3cccc(F)c3)cc1C(C)=O)C2=O.[Br-]', 'CN(C)C(=O)Oc1cccc2c1cc(C(=O)NC1CCCCC1)c[n+]2C.[I-]', 'COc1ccc(C(=O)CCC2C=CN(Cc3cccc(Cl)c3)C=C2C#N)cc1OC', 'O=C(O)C(F)(F)F.O=C1C2=C(C3=NCCc4c[nH]c1c43)C1(C=CN2)C=C(Br)C(O)C(Br)=C1', 'COc1cc2c(cc1OC)C(=O)C(CC1C=CN(Cc3ccccc3)C=C1C(C)=O)CC2', 'COc1cc2c(cc1OC)C(=O)C(CC1C=CN(Cc3ccccc3)C=C1C(C)=O)CC2', 'COc1cc2c(cc1OC)C(=O)C(CC1C=CN(Cc3ccccc3)C=C1C(C)=O)CC2', 'COc1cc2c(cc1OC)C(=O)C(CC1C=CN(Cc3ccccc3)C=C1C(C)=O)CC2', 'O=C(Nc1cc(Cl)nc(Cl)c1)OCCCCCCNCc1ccccc1', 'COc1ccccc1C1NC(=S)Nc2nc3c(c(N)c21)CCCC3', 'COc1ccc2c(c1)CCC(Cc1cc[n+](Cc3ccccc3C)cc1C(C)=O)C2=O.[Br-]', 'O=C(NCCCCCCNCc1ccccc1)c1cc(Cl)nc(Cl)c1', 'COc1cc2c(cc1OC)C(=O)C(Cc1cc[n+](Cc3ccccc3)cc1C(C)=O)CC2.[Br-]', 'O=C1SCc2ccccc21', 'COc1cc2c(cc1OC)C(=O)C(Cc1cc[n+](Cc3nc4ccccc4o3)cc1C(N)=O)C2.[Br-]', 'Cc1ccc(C2NC(=S)Nc3nc4c(c(N)c32)CCCC4)cc1', 'COc1ccc(C(=O)CCc2cc[n+](Cc3cccc(F)c3)cc2C#N)cc1OC.[Br-]', 'CC1(c2nccc3cc4c(cc23)OCO4)OC(O)c2c1ccc1c2OCO1', 'CC1(c2nccc3cc4c(cc23)OCO4)OC(O)c2c1ccc1c2OCO1', 'O=C(NCCCCCCNCc1ccc2ccccc2c1)c1cc(Cl)nc(Cl)c1', 'Cc1cc(=O)oc2cc(OCCCCCCSC(=S)N3CCCCC3)ccc12', 'O=C(Nc1cc(Cl)nc(Cl)c1)OCCCCCNCc1c[nH]c2ccccc12', 'Nc1c2c(nc3c1C(c1cccc([N+](=O)[O-])c1)NC(=S)N3)CCCC2', 'O=C(Nc1cc(Cl)nc(Cl)c1)OCCCCCNCc1ccccc1', 'Cc1cc(=O)oc2cc(OCCSC(=S)N3CCCC3)ccc12', 'CNC(=O)c1cc2c(OC(=O)N(C)C)cccc2[n+](C)c1.[I-]', 'COc1ccc2c(c1)CCC(Cc1cc[n+](Cc3ccccc3C)cc1C(F)(F)F)C2=O', 'CN(C)C(=O)Oc1cccc2c1cc(C(=O)NCCOCCO)c[n+]2C.COS(=O)(=O)[O-]', 'O=C(NCCCCCCNCc1c[nH]c2ccccc12)c1cc(Cl)nc(Cl)c1', 'COc1cc2c(cc1OC)C(=O)C(Cc1cc[n+](Cc3ccccc3)cc1Br)C2.[Br-]', 'c1cc2cc3c(cc2c([C@H]2Oc4ccc5c(c4O2)OCO5)n1)OCO3', 'COc1ccc(C(=O)CCc2cc[n+](Cc3cccc(Cl)c3)cc2C(C)=O)cc1OC.[Br-]', 'COc1ccc(C(=O)CCc2cc[n+](Cc3cccc(Cl)c3)cc2C(=O)O)cc1OC.[Br-]', 'Cc1cccc(C2NC(=S)Nc3nc4c(c(N)c32)CCCC4)c1', 'CNC(=O)c1cc2c(OC(=O)N(C)C)cccc2[n+](Cc2ccccc2)c1.[Br-]', 'COc1cc2c(cc1OC)C(=O)/C(=C/C1CCN(Cc3cc(N)c4ccccc4n3)CC1)C2', 'COc1ccc2c(c1)CCC(Cc1cc[n+](Cc3ccccc3Cl)cc1C(N)=O)C2=O.[Br-]', 'COc1ccc2c(c1)CCC(Cc1cc[n+](Cc3cccc(Cl)c3)cc1C(C)=O)C2=O.[Br-]', 'O=C1SCc2cccc([N+](=O)[O-])c21', 'Cc1cc(=O)oc2cc(OCCCCSC(=S)N3CCCCC3)ccc12', 'Cc1cc(=O)oc2cc(OCCCCSC(=S)N3CCCCC3)ccc12', 'Nc1c2c(nc3c1C(c1ccc(Cl)cc1)NC(=S)N3)CCCC2', 'COc1ccc2c3c1O[C@H]1C[C@H](O)CC[C@@]31CCN(C)C2', 'Cc1cc(=O)oc2cc(OCCSC(=S)N3CCN(C(C)C)CC3)ccc12', 'COc1ccc2c(c1)CCC(Cc1cc[n+](Cc3ccccc3F)cc1C(C)=O)C2=O.[Br-]', 'Nc1c2c(nc3c1C(c1ccc4c(c1)OCO4)NC(=S)N3)CCCC2', 'COC(=O)CNC(=O)c1cc2c(OC(=O)N(C)C)cccc2[n+](C)c1.[I-]', 'CO[C@H]1O[C@@](C)(c2nccc3cc4c(cc23)OCO4)c2ccc3c(c21)OCO3', 'CN(C)C(=O)Oc1cccc2c1cc(C(=O)NCCOCCOCCN1C(=O)c3ccccc3C1=O)c[n+]2C.O=S(=O)([O-])C(F)(F)F', 'CN(C)C(=O)Oc1cccc2c1cc(C(=O)Nc1ccccc1)c[n+]2C.[I-]', 'O=C(O)C(F)(F)F.O=C1CC[C@@]2(C=CNC3=C2C2=NCCc4c[nH]c(c42)C3=O)C=C1Br', 'COc1ccccc1CNCCCCCOC(=O)Nc1cc(Cl)nc(Cl)c1', 'COc1ccc2c(c1)CCC(Cc1cc[n+](Cc3cccc(C)c3)cc1C(C)=O)C2=O.[Br-]', 'Cc1cc(=O)oc2cc(OCCSC(=S)N(C)C)ccc12', 'COc1cc2c(cc1OC)C(=O)C(Cc1cc[n+](Cc3cccc(Cl)c3)cc1C(C)=O)CC2.[Br-]', 'CC(=O)c1c[n+](Cc2cccc(Cl)c2)ccc1CC1Cc2cc(O)ccc2C1=O.[Br-]', 'COc1cc2c(cc1OC)C(=O)C(Cc1cc[n+](Cc3cccc(C)n3)cc1C(N)=O)C2.[Br-]', 'Nc1c2c(nc3c1C(c1ccccc1)NC(=S)N3)CCCC2', 'COc1cccc(C2NC(=S)Nc3nc4c(c(N)c32)CCCC4)c1', 'COc1cccc(C2NC(=S)Nc3nc4c(c(N)c32)CCCC4)c1', 'COC(=O)c1c[n+](Cc2cccc(Cl)c2)ccc1CCC(=O)c1ccc(OC)c(OC)c1.[Br-]', 'COC1c2c(ccc3c2OCO3)-c2ccc3cc4c(cc3c2N1C)OCO4', 'CC(C)NC(=O)c1cc2c(OC(=O)N(C)C)cccc2[n+](C)c1.[I-]', 'COc1cc2c3c(c1)c1cc(=O)c(OC)cc-1cn3CC2', 'COc1ccc2c(c1)CCC(Cc1cc[n+](Cc3ccc(Cl)cc3)cc1C(C)=O)C2=O.[Br-]', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN[C@@H]1N2.O=C(O)c1ccccc1O', 'CNC(=O)Oc1ccc2c(c1)[C@]1(C)CCN[C@@H]1N2.O=C(O)c1ccccc1O', 'O=C(Nc1ccncc1Br)OCCCCCNCc1ccccc1', 'COc1cc2c(cc1OC)C(=O)/C(=C/C1CCN(Cc3cc(N=[N+]=[N-])c4ccccc4n3)CC1)C2', 'COc1ccc2c(c1)CCC(Cc1cc[n+](Cc3ccc(C)cc3)cc1C(C)=O)C2=O.[Br-]', 'COc1ccc2c(c1)CCC(Cc1cc[n+](Cc3cccc(C)c3)cc1Br)C2=O.[Br-]', 'C#CCN(C)Cc1cc2cc(OCCCN3CCCCC3)ccc2n1C', 'C#CCN(C)Cc1cc2cc(OCCCN3CCCCC3)ccc2n1C', 'C#CCN(C)Cc1cc2cc(OCCCN3CCCCC3)ccc2n1C', 'C#CCN(C)Cc1cc2cc(OCCCN3CCCCC3)ccc2n1C', 'COc1cc2c(cc1OC)C(=O)C(Cc1cc[n+](CCc3ccccc3)cc1C(N)=O)C2.[Br-]', 'COc1ccc2c(c1)CCC(Cc1cc[n+](Cc3cccc(C(F)(F)F)c3)cc1C(C)=O)C2=O.[Br-]', 'CC(=O)c1c[n+](Cc2ccccc2C)ccc1CC1Cc2cc(OCC(F)(F)F)ccc2C1=O.[Br-]', 'COc1cc2c(cc1OC)C(=O)C(CC1CCN(Cc3cc(N)c4ccccc4n3)CC1)C2', 'c1ccc2c(c1)CN[C@@H](CN(CC[C@H]1CCCNC1)[C@H]1CCCc3cccnc31)C2', 'COc1ccc2c(c1)CCC(Cc1cc[n+](Cc3ccccc3C)cc1C(N)=O)C2=O.[Br-]', 'Cc1cc(=O)oc2cc(OCCCCCSC(=S)N3CCCCC3)ccc12', 'Nc1c2c(nc3c1C(c1cccc(Br)c1)NC(=S)N3)CCCC2', 'Nc1c2c(nc3c1C(c1cccc(Br)c1)NC(=S)N3)CCCC2', 'COc1ccc(C(=O)CCc2cc[n+](Cc3ccccc3C(F)(F)F)cc2C#N)cc1OC.[Br-]', 'Cc1cc(=O)oc2cc(OCCSC(=S)N3CCCCC3)ccc12', 'CC[n+]1cc(C(=O)NC)cc2c(OC(=O)N(C)C)cccc21.[I-]', 'COc1ccccc1CNCCCCCCNC(=O)c1cc(Cl)nc(Cl)c1', 'CN(C)c1ccc(C2NC(=S)Nc3nc4c(c(N)c32)CCCC4)cc1', 'COc1ccc2c(c1)CCC(Cc1cc[n+](CC3CCCCC3)cc1C(C)=O)C2=O.[I-]', 'COCCNC(=O)c1cc2c(OC(=O)N(C)C)cccc2[n+](C)c1.[I-]', 'COc1ccc(C2NC(=S)Nc3nc4c(c(N)c32)CCCC4)cc1', 'COC(=O)c1cc2c(OC(=O)N(C)C)cccc2[n+](C)c1.[I-]', 'CN(C)C(=O)Oc1cccc2c1cc(C(=O)N(C)C)c[n+]2C.[I-]', 'COc1ccc2c(c1)CCC(Cc1cc[n+](Cc3ccccc3Cl)cc1C(C)=O)C2=O.[Br-]', 'CCN(CC)C(=S)SCCOc1ccc2c(C)cc(=O)oc2c1', 'N#Cc1ccccc1CNCC(=O)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'COc1ccc(Cn2cc(C(=O)NCCCCCNc3c4c(nc5ccccc35)CCCC4)c(=O)c3ccccc32)cc1', 'COc1ccc(CC(=O)Nc2ccc(CN(C)Cc3ccccc3)cc2)cc1OC', 'CCN(Cc1ccccc1)Cc1ccc(NC(=O)Cc2ccc(OC)c(OC)c2)cc1', 'COc1ccc(Cn2cc(C(=O)NCCCCCCCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)c(=O)c3ccccc32)cc1', 'CCCCCCCCNCC(=O)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'c1ccc2c(NCCCCC[Se][Se]CCCCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'COc1ccc(Cn2cc(C(=O)NCCCCCCCCNc3c4c(nc5ccccc35)CCCC4)c(=O)c3ccccc32)cc1', 'COc1ccc(Cn2cc(C(=O)NCCCCCNc3c4c(nc5ccc(OC)cc35)CCCC4)c(=O)c3ccccc32)cc1', 'Cc1cc(=O)oc2cc(OCc3cc[n+](Cc4cccc([N+](=O)[O-])c4)cc3)ccc12.[Br-]', 'COc1ccccc1C(=O)Nc1ccc2c(c1)CN(C(=O)c1cccc(C)c1)C(=O)C2', 'COc1ccc(Cn2cc(C(=O)NCCNc3c4c(nc5ccccc35)CCCC4)c(=O)c3ccccc32)cc1', 'COc1ccccc1C(=O)N1Cc2cc(NC(=O)c3cccc(Cl)c3)ccc2CC1=O', 'CCCCCCCCCCCCCCCCNCC(=O)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'Cc1cccc(C[n+]2ccc(COc3ccc4c(C)cc(=O)oc4c3)cc2)c1.[Br-]', 'O=C(CNCc1ccccc1F)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'Cc1cc(=O)oc2cc(OCc3cc[n+](Cc4ccccc4F)cc3)ccc12.[Br-]', 'COc1ccc(Cn2cc(C(=O)NCCCCCCNc3c4c(nc5ccc(OC)cc35)CCCC4)c(=O)c3ccccc32)cc1', 'O=C(Nc1ccc2c(c1)CN(C(=O)c1ccccc1)C(=O)C2)c1cccc(Cl)c1', 'Cc1cc(=O)oc2cc(OCc3cc[n+](Cc4cccc(F)c4)cc3)ccc12.[Br-]', 'COc1ccc(Cn2cc(C(=O)NCCCCCCNc3c4c(nc5ccccc35)CCCC4)c(=O)c3ccccc32)cc1', 'O=C(CNCc1cc(F)cc(F)c1)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'O=C(CNCc1cc(F)cc(F)c1)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'O=C(CNCc1cc(F)cc(F)c1)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'CC(C)=CCNCC(=O)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'CCN(Cc1ccccc1)Cc1ccc(NC(=O)Cc2ccc(O)cc2)cc1', 'CCCCNCC(=O)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'CN(Cc1ccccc1)Cc1ccc(NC(=O)Cc2ccc(O)cc2)cc1', 'c1ccc2c(NCCCCCSSCCCCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'O=C(Nc1ccc2c(c1)CN(C(=O)c1cccc(Cl)c1)C(=O)C2)c1ccccc1', 'CCCCCCCCCCNCC(=O)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'Cc1cc(=O)oc2cc(OCc3cc[n+](Cc4ccccc4Br)cc3)ccc12.[Br-]', 'CCNCC(=O)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'COc1ccccc1C(=O)Nc1ccc2c(c1)CN(C(=O)c1ccccc1)C(=O)C2', 'COc1ccc(Cn2cc(C(=O)NCCCCNc3c4c(nc5ccccc35)CCCC4)c(=O)c3ccccc32)cc1', 'COc1cc(/C=C/C(=O)NCCCCCCNc2c3c(nc4ccccc24)CCCC3)cc(OC)c1OCc1nc(C)c(C)nc1C', 'Cc1cccc(C(=O)Nc2ccc3c(c2)CN(C(=O)c2cccc(C)c2)C(=O)C3)c1', 'O=C(CNCCCc1ccccc1)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'COc1ccccc1C(=O)Nc1ccc2c(c1)CN(C(=O)c1ccccc1OC)C(=O)C2', 'COc1ccc(Cn2cc(C(=O)NCCCCCCCCNc3c4c(nc5ccc(OC)cc35)CCCC4)c(=O)c3ccccc32)cc1', 'Cc1cccc(C(=O)N2Cc3cc(NC(=O)c4cccc(Cl)c4)ccc3CC2=O)c1', 'Cc1ccccc1C[n+]1ccc(COc2ccc3c(C)cc(=O)oc3c2)cc1.[Br-]', 'Cc1ccccc1CNCC(=O)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'COc1ccc(Cn2cc(C(=O)NCCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)c(=O)c3ccccc32)cc1', 'COc1ccc(Cn2cc(C(=O)NCCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)c(=O)c3ccccc32)cc1', 'COc1ccc(Cn2cc(C(=O)NCCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)c(=O)c3ccccc32)cc1', 'CC(=O)OC(C)(C)/C=C/C(=O)[C@](C)(O)[C@H]1[C@H](O)C[C@@]2(C)[C@@H]3CC=C(C(C)(C)C(=O)O)/C(=C\\C(=O)O)[C@]3(C)C(=O)C[C@]12C', 'COc1ccc(Cn2cc(C(=O)NCCNc3c4c(nc5ccc(OC)cc35)CCCC4)c(=O)c3ccccc32)cc1', 'CCN(Cc1ccccc1)Cc1ccc(NC(=O)Cc2ccc(O)c(O)c2)cc1', 'C#CCNCC(=O)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'C=CCNCC(=O)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'c1ccc2c(NCCCCC[Se]CCCCCNc3c4c(nc5ccccc35)CCCC4)c3c(nc2c1)CCCC3', 'CCCNCC(=O)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'COc1ccccc1C(=O)N1Cc2cc(NC(=O)c3ccccc3)ccc2CC1=O', 'Cc1cc(=O)oc2cc(OCc3cc[n+](Cc4ccc(Br)cc4)cc3)ccc12.[Br-]', 'COc1ccc(Cn2cc(C(=O)NCCCCCCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)c(=O)c3ccccc32)cc1', 'O=C(Nc1ccc2c(c1)CN(C(=O)c1ccccc1)C(=O)C2)c1ccccc1', 'Cc1ccc(C[n+]2ccc(COc3ccc4c(C)cc(=O)oc4c3)cc2)cc1.[Br-]', 'CC(=O)OC(C)(C)/C=C/C(=O)[C@](C)(O)[C@H]1[C@H](O)C[C@@]2(C)[C@@H]3CC=C(C(C)(C)C(=O)O)C4=CC(=O)O[C@](O)(C[C@]12C)[C@@]43C', 'COc1ccccc1C(=O)Nc1ccc2c(c1)CN(C(=O)c1cccc(Cl)c1)C(=O)C2', 'COc1ccc(Cn2cc(C(=O)NCCCNc3c4c(nc5ccccc35)CCCC4)c(=O)c3ccccc32)cc1', 'Cc1cccc(C(=O)Nc2ccc3c(c2)CN(C(=O)c2cccc(Cl)c2)C(=O)C3)c1', 'CCN(Cc1ccccc1)Cc1ccc(NC(=O)Cc2ccc(OC)c(O)c2)cc1', 'COc1ccc(Cn2cc(C(=O)NCCCCCCCNc3c4c(nc5ccccc35)CCCC4)c(=O)c3ccccc32)cc1', 'O=C(CNCc1ccccc1)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'CN(Cc1ccccc1)Cc1ccc(NC(=O)Cc2ccc(O)c(O)c2)cc1', 'COc1ccc(Cn2cc(C(=O)NCCCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)c(=O)c3ccccc32)cc1', 'COc1ccc(Cn2cc(C(=O)NCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)c(=O)c3ccccc32)cc1', 'COc1cc(/C=C/C(=O)NCCCCCCNc2c3c(nc4ccccc24)CCCC3)cc(OC)c1O', 'COc1ccc(Cn2cc(C(=O)NCCCCNc3c4c(nc5ccc(OC)cc35)CCCC4)c(=O)c3ccccc32)cc1', 'O/N=C/c1nc(CCCCNc2c3c(nc4ccc(Cl)cc24)CCCC3)ccc1O', 'O/N=C/c1nc(CCCCNc2c3c(nc4ccc(Cl)cc24)CCCC3)ccc1O', 'O/N=C/c1nc(CCCCNc2c3c(nc4ccc(Cl)cc24)CCCC3)ccc1O', 'O/N=C/c1nc(CCCCNc2c3c(nc4ccc(Cl)cc24)CCCC3)ccc1O', 'O/N=C/c1nc(CCCCNc2c3c(nc4ccc(Cl)cc24)CCCC3)ccc1O', 'CN(Cc1ccccc1)Cc1ccc(NC(=O)/C=C/c2ccc(O)c(O)c2)cc1', 'Cc1cc(=O)oc2cc(OCc3cc[n+](Cc4ccc([N+](=O)[O-])cc4)cc3)ccc12.[Br-]', 'COc1ccc(Cn2cc(C(=O)NCCCNc3c4c(nc5ccc(OC)cc35)CCCC4)c(=O)c3ccccc32)cc1', 'CCCCCCCCCCCCNCC(=O)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'COc1ccc(CC(=O)Nc2ccc(CN(C)Cc3ccccc3)cc2)cc1O', 'COc1ccc(Cn2cc(C(=O)NCCCCCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)c(=O)c3ccccc32)cc1', 'Cc1cc(=O)oc2cc(OCc3cc[n+](Cc4ccccc4)cc3)ccc12.[Br-]', 'COc1ccc(Cn2cc(C(=O)NCCCCCCCNc3c4c(nc5ccc(OC)cc35)CCCC4)c(=O)c3ccccc32)cc1', 'COc1ccc(Cn2cc(C(=O)NCCCCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)c(=O)c3ccccc32)cc1', 'O=C(Nc1ccc2c(c1)CN(C(=O)c1cccc(Cl)c1)C(=O)C2)c1cccc(Cl)c1', 'C=CCCNCC(=O)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'CCCCCCNCC(=O)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'Cc1cc(=O)oc2cc(OCc3cc[n+](Cc4ccc(F)cc4)cc3)ccc12.[Br-]', 'Cc1cccc(C(=O)N2Cc3cc(NC(=O)c4ccccc4)ccc3CC2=O)c1', 'Cc1cc(=O)oc2cc(OCc3cc[n+](Cc4cccc(Br)c4)cc3)ccc12.[Br-]', 'O=C(CNCc1ccccc1[N+](=O)[O-])Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'CN(Cc1ccccc1)Cc1ccc(NC(=O)Cc2ccccc2)cc1', 'C#CCNc1nc(NCCCCC2CCN(Cc3ccccc3)CC2)c(C#N)cc1C#N', 'COc1ccc2c(c1)C/C(=C/c1ccccc1-c1ccc3nc(N)ccc3c1)C2=O', 'COc1ccc2c(c1)C/C(=C/C1CCN(Cc3ccccc3)CC1)C2=O', 'COc1ccc2c(c1)C/C(=C/c1ccccc1-c1ccc3ccccc3c1)C2=O', 'COc1ccc2c(c1)C/C(=C/c1ccccc1-c1ccccc1)C2=O', 'COc1ccc(-c2ccccc2/C=C2/Cc3cc(OC)ccc3C2=O)cc1', 'CN(CCCCCN1C(=O)c2ccccc2S1(=O)=O)Cc1ccccc1', 'CCN1CCN(C2CCN(C(=O)COc3ccc(-c4ccccc4F)cc3)CC2)CC1', 'CCN(CC)CCOc1ccc(/C=C2\\COc3ccccc3C2=O)cc1', 'O=C(NCCC1CCN(Cc2ccccc2)CC1)c1cc(=O)c2c(O)cc(O)cc2o1', 'O=C(COc1ccc(-c2ccccc2)cc1)NCCCN1CCCCC1', 'O=C(NCCC1CCN(Cc2ccccc2)CC1)c1cc(=O)c2cc([N+](=O)[O-])ccc2o1', 'O=C(NCCC1CCN(Cc2ccccc2)CC1)c1cc(=O)c2cc([N+](=O)[O-])ccc2o1', 'Cc1ccccc1CN1CCC(CCNC(=O)/C=C/c2cc(C(C)(C)C)c(O)c(C(C)(C)C)c2)CC1', 'Cc1[nH]c2ccccc2c1CC(=O)NCCC1CCN(Cc2ccccc2)CC1', 'NC(=O)c1cc[n+](C/C=C/C[n+]2cc(Cl)c(/C=N/O)c(Cl)c2)cc1.[Br-].[Br-]', 'O=C(COc1ccc(-c2ccccc2)cc1)N1CCC(N2CCCCC2)CC1', 'COC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCNc1c2c(nc3ccccc13)CCCC2', 'CC(C)(C)c1cc(/C=C/C(=O)NCCC2CCN(Cc3ccc(C#N)cc3)CC2)cc(C(C)(C)C)c1O', 'COc1ccc2nc3c(c(NCCCNC(=O)c4sc(-c5ccnc(NC(=O)C6CC6)c5)nc4OC)c2c1)CCCC3', 'CCN(CC)CCOc1cccc(/C=C2\\COc3ccccc3C2=O)c1', 'COc1nc(-c2ccnc(NC(=O)C3CC3)c2)sc1C(=O)NCCCNc1c2c(nc3ccccc13)CCCC2', 'CCN(CC)CCOC(=O)/C=C/c1ccc2occ(C(=O)Nc3ccc(C)cc3)c(=O)c2c1', 'O=C(O)/C=C/c1ccc2oc(C(=O)Nc3ccccc3)cc(=O)c2c1', 'Cc1ccc(NC(=O)c2cc(=O)c3cc(/C=C/C(=O)OCCN(C)C)ccc3o2)cc1C', 'CC(C)(C)c1cc(C(=O)CNCCC2CCN(Cc3ccccc3)CC2)cc(C(C)(C)C)c1O', 'CCN(CC)Cc1cc(N)ccc1O.Cl.Cl', 'CCN(CC)Cc1cc(N)ccc1O.Cl.Cl', 'CCN(CC)Cc1cc(N)ccc1O.Cl.Cl', 'CCN(CC)Cc1cc(N)ccc1O.Cl.Cl', 'CCN(CC)Cc1cc(N)ccc1O.Cl.Cl', 'CCN(CC)Cc1cc(N)ccc1O.Cl.Cl', 'Fc1ccc2c(C3CCN(CCN4CCCC4)CC3)noc2c1', 'Nc1c2c(nc3cc(Cl)ccc13)C[C@H]1C=C(CCCCn3cc(CO)nn3)C[C@H]2C1', 'O=C(NCCC1CCN(Cc2ccccc2)CC1)c1cc(=O)c2cc(O)c(O)cc2o1', 'CCN1CCN(CCCNC(=O)COc2ccc(I)cc2)CC1', 'CCN(CC)CCCCCCOc1ccc(/C=C2\\COc3ccccc3C2=O)cc1', 'CC(C)(C)c1cc(/C=C/C(=O)NCCC2CCN(Cc3ccc(F)c(F)c3)CC2)cc(C(C)(C)C)c1O', 'NC(=O)c1cc[n+](CCC[n+]2cc(Cl)c(/C=N/O)c(Cl)c2)cc1.[Br-].[Br-]', 'C[N+](C)(CC#CCOC1=NOCC1)CCCCCCCCCC[N+](C)(C)CCCN1C(=O)c2ccccc2C1=O.[Br-].[Br-]', 'Cn1c2nc3ccccc3c-2c(NCCCCNc2c3c(nc4ccccc24)CCCC3)c2ccccc21', 'CC(C)(C)c1cc(/C=C/C(=O)NCCC2CCN(Cc3ccc(Br)cc3)CC2)cc(C(C)(C)C)c1O', 'CN(C)CCOc1cccc(/C=C2\\COc3ccccc3C2=O)c1', 'CN(C)CCOC(=O)/C=C/c1ccc2oc(C(=O)Nc3ccc(Cl)cc3)cc(=O)c2c1', 'COc1nc(-c2ccnc(NC(=O)C3CC3)c2)sc1C(=O)NCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Cl.Cl.Nc1ccc(O)c(CN2CCCC2)c1', 'Cl.Cl.Nc1ccc(O)c(CN2CCCC2)c1', 'Cl.Cl.Nc1ccc(O)c(CN2CCCC2)c1', 'Cn1cc(C(=O)NCCC2CCN(Cc3ccccc3)CC2)c2ccccc21', 'COc1ccc2c(=O)cc(C(=O)NCCC3CCN(Cc4ccccc4)CC3)[nH]c2c1', 'CC(C)(C)c1cc(/C=C/C(=O)NCCC2CCN(Cc3cccc(Cl)c3)CC2)cc(C(C)(C)C)c1O', 'CCN(CC)CCOC(=O)/C=C/c1ccc2oc(C(=O)Nc3ccc(C)cc3)cc(=O)c2c1', 'O=C(NC1CCN(Cc2ccccc2)CC1)c1cc(=O)c2cc(O)ccc2o1', 'CCN(CC)CCOC(=O)/C=C/c1ccc2oc(C(=O)Nc3cccc(Cl)c3)cc(=O)c2c1', 'C[N+](C)(CC#CCOC1=NOCC1)CCCCCCCNc1c2c(nc3ccccc13)CCCC2.[Br-]', 'CC(C)(C)c1cc(/C=C/C(=O)NCCC2CCN(Cc3ccccc3)CC2)cc(C(C)(C)C)c1O', 'CC(C)(C)c1cc(/C=C/C(=O)NCCC2CCN(Cc3ccccc3)CC2)cc(C(C)(C)C)c1O', 'COc1ccc2oc(C(=O)NCCC3CCN(Cc4ccccc4)CC3)cc(=O)c2c1', 'CCN(CC)CCOC(=O)/C=C/c1ccc2occ(C(=O)Nc3cccc(Cl)c3)c(=O)c2c1', 'O=C(Cc1c[nH]c2ccccc12)NCCC1CCN(Cc2ccccc2)CC1', 'CCN(CCOc1ccc(/C=C2\\COc3ccccc3C2=O)cc1)Cc1ccccc1', 'CN(C)CCOc1ccc(/C=C2\\COc3ccccc3C2=O)cc1', 'Cc1ccc(NC(=O)c2cc(=O)c3cc(/C=C/C(=O)O)ccc3o2)cc1', 'Cn1c2nc3ccccc3c-2c(NCCCCCCNc2c3c(nc4ccccc24)CCCC3)c2ccccc21', 'O=C(NCCCC1CCN(Cc2ccccc2)CC1)c1cc(=O)c2cc(O)ccc2o1', 'O=C(NCCCC1CCN(Cc2ccccc2)CC1)c1cc(=O)c2cc(O)ccc2o1', 'Cc1ccc(NC(=O)c2coc3ccc(/C=C/C(=O)OCCN(C)C)cc3c2=O)cc1', 'CC(C)(C)c1cc(/C=C/C(=O)NCCC2CCN(Cc3ccc(F)cc3)CC2)cc(C(C)(C)C)c1O', 'COc1cc2oc(C(=O)OCCC3CCN(Cc4ccccc4)CC3)cc(=O)c2cc1OC', 'COc1cc2oc(C(=O)OCCC3CCN(Cc4ccccc4)CC3)cc(=O)c2cc1OC', 'O=C(COc1ccc(-c2ccccc2)cc1)N1CCC(N2CCOCC2)CC1', 'O=C(O)/C=C/c1ccc2occ(C(=O)Nc3ccccc3)c(=O)c2c1', 'COc1cc(OC)c2c(=O)cc(C(=O)NCCC3CCN(Cc4ccccc4)CC3)oc2c1', 'CCN1CCN(C2CCN(C(=O)COc3ccc(-c4ccc(OC)cc4)cc3)CC2)CC1', 'NC(=O)c1cc[n+](CCCC[n+]2ccc(/C=N/O)c(Cl)c2)cc1.[Br-].[Br-]', 'Cc1ccc(CN2CCC(CCNC(=O)/C=C/c3cc(C(C)(C)C)c(O)c(C(C)(C)C)c3)CC2)cc1', 'CCN(CC)CCCCOc1ccc(/C=C2\\COc3ccccc3C2=O)cc1', 'CCOC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCNc1c2c(nc3ccccc13)CCCC2', 'COc1ccc2nc3c(c(NCCNC(=O)c4sc(-c5ccnc(NC(=O)C6CC6)c5)nc4OC)c2c1)CCCC3', 'COc1cc2oc(C(=O)OCC3CCN(Cc4ccccc4)CC3)cc(=O)c2cc1OC', 'CN(C)CCOC(=O)/C=C/c1ccc2oc(C(=O)Nc3ccccc3)cc(=O)c2c1', 'CN(C)CCOC(=O)/C=C/c1ccc2occ(C(=O)Nc3ccc(Cl)cc3)c(=O)c2c1', 'Cc1cccc2c1ccc1c3c(c(=O)oc(=O)c12)[C@@H](C)CO3', 'COc1ccc2c(=O)cc(C(=O)NCCC3CCN(Cc4ccccc4)CC3)oc2c1', 'O=C(Cc1c[nH]c2ccc(Br)cc12)NCCC1CCN(Cc2ccccc2)CC1', 'NC(=O)c1cc[n+](CCCC[n+]2cc(Cl)c(/C=N/O)c(Cl)c2)cc1.[Br-].[Br-]', 'O=C(NCCC1CCN(Cc2ccccc2)CC1)c1cc(=O)c2cccc(O)c2[nH]1', 'CN1CCN(C2CCN(C(=O)COc3ccc(-c4ccccc4)cc3)CC2)CC1', 'CN(C)CCOC(=O)/C=C/c1ccc2occ(C(=O)Nc3ccccc3)c(=O)c2c1', 'COc1nc(-c2ccnc(NC(=O)C3CC3)c2)sc1C(=O)NCCNc1c2c(nc3ccccc13)CCCC2', 'COc1ccc2oc(C(=O)OCC3CCN(Cc4ccccc4)CC3)cc(=O)c2c1', 'CC(C)(C)c1cc(C(=O)CNCC2CCN(Cc3ccccc3)CC2)cc(C(C)(C)C)c1O', 'O=C(NCCC1CCN(Cc2ccccc2)CC1)c1cc(=O)c2ccccc2o1', 'O=C(NCCC1CCN(Cc2ccccc2)CC1)c1cc(=O)c2ccccc2o1', 'CCN(CC)CCOC(=O)/C=C/c1ccc2occ(C(=O)Nc3ccc(OC)cc3)c(=O)c2c1', 'CCN1CCN(C2CCN(C(=O)COc3ccc4c(c3)CCC4=O)CC2)CC1', 'COc1ccc2[nH]c(C(=O)NCCC3CCN(Cc4ccccc4)CC3)cc(=O)c2c1', 'COc1ccc(NC(=O)c2cc(=O)c3cc(/C=C/C(=O)OCCN(C)C)ccc3o2)cc1', 'C[N+](C)(CC#CCOC1=NOCC1)CCCCCCCCCCNc1c2c(nc3ccccc13)CCCC2.[Br-]', 'CC(C)(C)c1cc(CCC(=O)NCCC2CCN(Cc3ccccc3)CC2)cc(C(C)(C)C)c1O', 'COc1ccc2oc(C(=O)NCC3CCN(Cc4ccccc4)CC3)cc(=O)c2c1', 'COc1ccc(NC(=O)c2coc3ccc(/C=C/C(=O)O)cc3c2=O)cc1', 'COc1cc2oc(C(=O)NCCC3CCN(Cc4ccccc4)CC3)cc(=O)c2cc1OC', 'COc1cc2oc(C(=O)NCCC3CCN(Cc4ccccc4)CC3)cc(=O)c2cc1OC', 'COc1ccc(CN2CCC(CCNC(=O)/C=C/c3cc(C(C)(C)C)c(O)c(C(C)(C)C)c3)CC2)cc1', 'O=C1/C(=C/c2ccc(OCCCN3CCN(Cc4ccccc4)CC3)cc2)COc2ccccc21', 'NC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(COc1ccc(I)cc1)NCCCN1CCCCC1', 'CN(C)CCCOc1cccc(/C=C2\\COc3ccccc3C2=O)c1', 'NC(=O)c1cc[n+](C/C=C/C[n+]2ccc(/C=N/O)c(Cl)c2)cc1.[Br-].[Br-]', 'COc1cc2nc(C(=O)NCCC3CCN(Cc4ccccc4)CC3)cc(OC)c2cc1OC', 'C[C@@H]1C[C@H]2CC(=O)[C@H]3CCC[N+]4([O-])CCC[C@]2(O)[C@]34C1', 'CN(C)CCOC(=O)/C=C/c1ccc2oc(C(=O)Nc3cccc(Cl)c3)cc(=O)c2c1', 'CCN(CC)CCOC(=O)/C=C/c1ccc2oc(C(=O)Nc3ccccc3)cc(=O)c2c1', 'COc1nc(-c2ccnc(NC(=O)C3CC3)c2)sc1C(=O)NCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'Cc1coc2c1c(=O)oc(=O)c1c3c(ccc12)C(C)(C)CCC3', 'COc1cccc2c(=O)cc(C(=O)NCCC3CCN(Cc4ccccc4)CC3)[nH]c12', 'Fc1ccc2c(C3CCN(CCN4CCCCC4)CC3)noc2c1', 'COc1ccc2oc(C(=O)NC3CCN(Cc4ccccc4)CC3)cc(=O)c2c1', 'COc1ccc(-c2ccc(OCC(=O)NCCCN3CCCCC3)cc2)cc1', 'CCN(CC)CCOC(=O)/C=C/c1ccc2oc(C(=O)Nc3ccc(Cl)cc3)cc(=O)c2c1', 'O=C(NCCC1CCN(Cc2ccccc2)CC1)c1cc(=O)c2cc(O)ccc2o1', 'COc1ccc2oc(C(=O)NCCCC3CCN(Cc4ccccc4)CC3)cc(=O)c2c1', 'CCN1CCN(C2CCN(C(=O)COc3ccc(I)cc3)CC2)CC1', 'CN(CCCCCCOc1ccc(/C=C2\\COc3ccccc3C2=O)cc1)Cc1ccccc1', 'CC(C)(C)c1cc(/C=C/C(=O)NCCC2CCN(Cc3ccc(F)cc3F)CC2)cc(C(C)(C)C)c1O', 'CCN1CCN(C2CCN(C(=O)COc3ccc(-c4ccc([N+](=O)[O-])cc4)cc3)CC2)CC1', 'CCN(CC)c1ccc(/C=C/c2cc(N3CCCCC3)c3ccccc3n2)cc1', 'Cc1ccc(NC(=O)c2cc(=O)c3cc(/C=C/C(=O)OCCN(C)C)ccc3o2)cc1', 'Cc1ccc(NC(=O)c2cc(=O)c3cc(/C=C/C(=O)OCCN(C)C)ccc3o2)cc1', 'CC(C)(C)c1cc(/C=C/C(=O)NCCC2CCN(Cc3cccc([N+](=O)[O-])c3)CC2)cc(C(C)(C)C)c1O', 'COc1cc2oc(C(=O)NC3CCN(Cc4ccccc4)CC3)cc(=O)c2cc1OC', 'Nc1ccc2oc(C(=O)NCCC3CCN(Cc4ccccc4)CC3)cc(=O)c2c1', 'Nc1ccc2oc(C(=O)NCCC3CCN(Cc4ccccc4)CC3)cc(=O)c2c1', 'NC(=O)c1cc[n+](CCC[n+]2ccc(/C=N/O)c(Cl)c2)cc1.[Br-].[Br-]', 'CC(C)(C)c1cc(/C=C/C(=O)NCCC2CCN(Cc3ccccc3F)CC2)cc(C(C)(C)C)c1O', 'CC(C)(C)c1cc(/C=C/C(=O)NCCC2CCN(Cc3ccccc3F)CC2)cc(C(C)(C)C)c1O', 'Cn1cc(CC(=O)NCCC2CCN(Cc3ccccc3)CC2)c2ccccc21', 'CC(C)(CN1C(=O)c2cccc3cccc(c23)C1=O)C[N+](C)(C)CCCCCCCCCC[N+](C)(C)CC#CCOC1=NOCC1.[Br-].[Br-]', 'CC(C)(C)c1cc(/C=C/C(=O)NCCC2CCN(Cc3ccccc3Cl)CC2)cc(C(C)(C)C)c1O', 'O=C(O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCNc1c2c(nc3ccccc13)CCCC2', 'CCN(CC)CCOC(=O)/C=C/c1ccc2oc(C(=O)Nc3ccc(OC)cc3)cc(=O)c2c1', 'COc1nc(-c2ccnc(NC(=O)C3CC3)c2)sc1C(=O)NCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'C[C@H]1COc2c1c(=O)oc(=O)c1c3c(ccc21)C(C)(C)CCC3', 'COc1cc2[nH]c(C(=O)NCCC3CCN(Cc4ccccc4)CC3)cc(=O)c2cc1OC', 'COc1cc2[nH]c(C(=O)NCCC3CCN(Cc4ccccc4)CC3)cc(=O)c2cc1OC', 'CCN(CC)CCOC(=O)/C=C/c1ccc2occ(C(=O)Nc3ccc(Cl)cc3)c(=O)c2c1', 'O=C(CCCc1c[nH]c2ccccc12)NCCC1CCN(Cc2ccccc2)CC1', 'COc1ccc(NC(=O)c2cc(=O)c3cc(/C=C/C(=O)O)ccc3o2)cc1', 'CCN(CC)CCOC(=O)/C=C/c1ccc2oc(C(=O)Nc3ccc(C)c(C)c3)cc(=O)c2c1', 'O=C(NCCC1CCN(Cc2ccccc2)CC1)c1cc(=O)c2cc(O)ccc2[nH]1', 'Cn1c2nc3ccccc3c-2c(NCCCCCCCCNc2c3c(nc4ccccc24)CCCC3)c2ccccc21', 'COc1cccc2oc(C(=O)NCCC3CCN(Cc4ccccc4)CC3)cc(=O)c12', 'COc1cc2oc(C(=O)NCCCC3CCN(Cc4ccccc4)CC3)cc(=O)c2cc1OC', 'COc1cc2oc(C(=O)NCCCC3CCN(Cc4ccccc4)CC3)cc(=O)c2cc1OC', 'COc1ccc(NC(=O)c2coc3ccc(/C=C/C(=O)OCCN(C)C)cc3c2=O)cc1', 'CC(C)(C)c1cc(/C=C/C(=O)NCCC2CCN(Cc3cccc(F)c3)CC2)cc(C(C)(C)C)c1O', 'COc1cc2oc(C(=O)NCC3CCN(Cc4ccccc4)CC3)cc(=O)c2cc1OC', 'S=C(NNc1c2c(nc3ccccc13)CCCC2)Nc1ccc(Cl)cc1', 'S=C(NNc1c2c(nc3ccccc13)CCCC2)Nc1ccc(Br)cc1', 'COc1cccc(CNC2CCN(Cc3ccccc3)CC2)c1OC', 'COc1ccc(C(=O)CNc2ccc3oc(=O)ccc3c2)cc1', 'CSc1ccc(/C=N/c2ccc3oc(=O)ccc3c2)cc1', 'Cl.Cl.N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCCNc1c2c(nc3ccccc13)CCCC2', 'O=C(CCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)NC12CC3CC(F)(CC(C1)c1ccccc13)C2', 'COc1cc2c(cc1OC)C(=O)C(Cc1cn(CC3CCN(Cc4ccccc4)CC3)nn1)C2', 'COc1ccc(-c2cc(=O)c3cc(OC)c(O)c([C@H]4CCC(=O)N4)c3o2)cc1', 'Oc1ccc(/C=N/c2c3c(nc4ccccc24)CCCC3)cc1', 'c1ccc(CN2CCC(Cn3cc(CNc4ncnc5ccccc45)nn3)CC2)cc1', 'COc1cc(C(=O)n2nc(OCCN3CCCC3)c3ccccc32)cc(OC)c1OC', 'Cl.O=[N+]([O-])c1ccc(-n2cc(COc3ccc(/C=N/Nc4c5c(nc6ccccc46)CCCC5)cc3)nn2)cc1', 'CN(C)c1ccc(/C=N/c2ccc3oc(=O)ccc3c2)cc1', 'CCOc1ccc(Cc2nc3cc(C(=O)NCCCNc4c5c(nc6ccccc46)CCCC5)ccc3n2CCC(C)C)cc1', 'CCOc1ccc(Cc2nc3cc(C(=O)NCCCNc4c5c(nc6ccccc46)CCCC5)ccc3n2CCC(C)C)cc1', 'CCOc1ccc(Cc2nc3cc(C(=O)NCCCNc4c5c(nc6ccccc46)CCCC5)ccc3n2CCC(C)C)cc1', 'COc1ccc2[nH]c3nc4c(c(N)c3c2c1)CCCC4', 'O=[N+]([O-])c1ccc(CNC2CCN(Cc3ccccc3)CC2)cc1', 'Oc1ccc(CNC2CCN(Cc3ccccc3)CC2)c(O)c1', 'O=C(Nc1cc(-c2cn(CCNc3c4c(nc5ccccc35)CCCC4)nn2)ccn1)Nc1cccc2c1[C@@H]1CCCCN1C2=O', 'Fc1ccc(CNC2CCN(Cc3ccccc3)CC2)c(F)c1', 'Cl.Clc1ccc(-n2cc(COc3ccc(/C=N/Nc4c5c(nc6ccccc46)CCCC5)cc3)nn2)cc1', 'O=C(Nc1cc(-c2cn(CCNc3c4c(nc5ccccc35)CCCC4)nn2)ccn1)Nc1cccc2c1[C@H]1CCCCN1C2=O', 'Cl.NS(=O)(=O)c1ccc(-n2cc(COc3ccc(/C=N/Nc4c5c(nc6ccccc46)CCCC5)cc3)nn2)cc1', 'COc1cc(NCc2ccc(Oc3c(N)cccc3Cl)cc2)cc(OC)c1OC', 'COc1cc(NCc2ccc(Oc3c(N)cccc3Cl)cc2)cc(OC)c1OC', 'O=C1CSC(c2ccc(F)cc2)N1c1ccc2oc(=O)ccc2c1', 'NS(=O)(=O)c1ccc(C(=O)n2[nH]c(=O)c3cc(F)ccc3c2=O)cc1', 'Cl.Cl.N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(CCC1CCN(CC2CCCCC2)CC1)c1cc(Cl)c2[nH]ccc2c1', 'Cc1ccc(CNC2CCN(Cc3ccccc3)CC2)cc1', 'C(=N/Nc1c2c(nc3ccccc13)CCCC2)\\c1ccc(OCc2cn(-c3ccccc3)nn2)cc1.Cl', 'OC(CNCc1cccc(OCc2ccccc2)c1)Cn1cc(Br)c2cc(Br)ccc21', 'Cc1cc(=O)oc2ccc(OC/C=C/CCn3cnc4c(N5CCCCC5)ncnc43)cc12', 'COc1ccc(C2=NN(C(C)=O)C(c3ccc(N4CCC(C)CC4)cc3)C2)cc1OC', 'O=C(c1cccc(Cl)c1Cl)n1nc(OCCN2CCCCC2)c2ccccc21', 'O=C(CNc1ccc2oc(=O)ccc2c1)c1ccccc1', 'COc1ccc(NCC(=O)Nc2ccc3oc(=O)ccc3c2)cc1', 'CN(Cc1ccnc(NC(=O)Nc2cccc3c2C2CCCCN2C3=O)c1)Cc1cn(CCNc2c3c(nc4ccccc24)CCCC3)nn1', 'c1ccc(CN2CCC(Cn3cc(CSc4nc5ccccc5[nH]4)nn3)CC2)cc1', 'CCOc1ccc(Cc2nc3cc(C(=O)NCCSSCCNc4c5c(nc6ccccc46)CCCC5)ccc3n2CCC(C)C)cc1', 'CCOc1ccc(Cc2nc3cc(C(=O)NCCSSCCNc4c5c(nc6ccccc46)CCCC5)ccc3n2CCC(C)C)cc1', 'CCOc1ccc(Cc2nc3cc(C(=O)NCCSSCCNc4c5c(nc6ccccc46)CCCC5)ccc3n2CCC(C)C)cc1', 'Cl.O=[N+]([O-])c1cccc(-n2cc(COc3ccc(/C=N/Nc4c5c(nc6ccccc46)CCCC5)cc3)nn2)c1', 'Oc1ccccc1CNC1CCN(Cc2ccccc2)CC1', 'O=C(CCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)NCCc1cc2ccccc2[nH]1', 'Cl.Cl.N[C@H](Cc1c[nH]c2ccccc12)C(=O)NCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=c1ccc2cc(/N=C/c3ccc(Cl)cc3)ccc2o1', 'FC(F)(F)Oc1ccc(/C=N/C2CCN(Cc3ccccc3)CC2)cc1', 'CSc1ccc(/C=C2/S/C(=N/c3ccc4oc(=O)ccc4c3)NC2=O)cc1', 'Cc1c(C)c2ccc(OCC3CCN(Cc4cccc(CO)c4)CC3)cc2oc1=O', 'COc1cc2oc(-c3ccc(NCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)cc3)cc(=O)c2c(OC)c1OC', 'O=C1N/C(=N\\c2ccc3oc(=O)ccc3c2)S/C1=C/c1ccccc1', 'O=[N+]([O-])c1ccc(/C=N/C2CCN(Cc3ccccc3)CC2)cc1', 'c1ccc(CNC2CCN(Cc3ccccc3)CC2)cc1', 'COc1ccc(NCc2ccc(Oc3c(N)cccc3Cl)cc2)cc1', 'NS(=O)(=O)c1ccc(C(=O)n2[nH]c(=O)c3cc([N+](=O)[O-])ccc3c2=O)cc1', 'Cc1c(C)c2ccc(OCC3CCCN(CCCO)C3)cc2oc1=O', 'CCOc1ccc(Cc2nc3cc(C(=O)NCCCCCNc4c5c(nc6ccccc46)CCCC5)ccc3n2CCC(C)C)cc1', 'Cc1cc(C)c(S(=O)(=O)Nc2ccc(C3C4=C(CC(C)(C)CC4=O)N(c4ccc(Br)cc4)C4=C3C(=O)CC(C)(C)C4)cc2)c(C)c1', 'Cc1cc(C)c(S(=O)(=O)Nc2ccc(C3C4=C(CC(C)(C)CC4=O)N(c4ccc(Br)cc4)C4=C3C(=O)CC(C)(C)C4)cc2)c(C)c1', 'Oc1ccc(/C=N/C2CCN(Cc3ccccc3)CC2)cc1', 'COc1c(C(=O)CCC2CCN(CC3CCCCC3)CC2)ccc2[nH]ccc12', 'COc1ccc(CNC(=O)C2CCN(C(=O)c3ccccc3)CC2)cc1OC', 'Cc1ccc2oc(-c3ccc(OCCCCCCN4CCCCCC4)cc3)cc(=O)c2c1', 'CCOc1ccc(Cc2nc3cc(C(=O)N(CC)CC)ccc3n2CCCCCCNc2c3c(nc4ccccc24)CCCC3)cc1', 'CCOc1ccc(Cc2nc3cc(C(=O)N(CC)CC)ccc3n2CCCCCCNc2c3c(nc4ccccc24)CCCC3)cc1', 'CCOc1ccc(Cc2nc3cc(C(=O)N(CC)CC)ccc3n2CCCCCCNc2c3c(nc4ccccc24)CCCC3)cc1', 'O=C(O)CC1(CC(=O)c2cc(O)ccc2O)C=C(CO)CCC1', 'CCOc1ccc(Cc2nc3cc(C(=O)N(CC)CC)ccc3n2CCOCCOCCNc2c3c(nc4ccccc24)CCCC3)cc1', 'Cl.Cl.N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'Cc1c(C)c2ccc(OCC3CCCN(Cc4ccccc4)C3)cc2oc1=O.Cl', 'c1ccc2c(NCc3ccc(-c4ccc(CNc5c6c(nc7ccccc57)CCCC6)cc4)cc3)c3c(nc2c1)CCCC3', 'CCOc1ccc(Cc2nc3cc(C(=O)N(CC)CC)ccc3n2CCNc2c3c(nc4ccccc24)CCCC3)cc1', 'S=C(NCc1ccccc1)NNc1c2c(nc3ccccc13)CCCC2', 'S=C(NCc1ccccc1)NNc1c2c(nc3ccccc13)CCCC2', 'S=C(NCc1ccccc1)NNc1c2c(nc3ccccc13)CCCC2', 'Cc1cccc(CNC2CCN(Cc3ccccc3)CC2)c1', 'CCOc1ccc(Cc2nc3cc(C(=O)N(CC)CC)ccc3n2CCCNc2c3c(nc4ccccc24)CCCC3)cc1', 'NS(=O)(=O)c1ccc(NC(=S)NNc2c3c(nc4ccccc24)CCCC3)cc1', 'Cl.Cl.N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCCCNc1c2c(nc3ccccc13)CCCC2', 'Cc1c(C)c2ccc(OCC3CCN(Cc4cccc(CO[N+](=O)[O-])c4)CC3)cc2oc1=O', 'Cl.Cl.N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCNc1c2c(nc3ccccc13)CCCC2', 'COc1ccc(-c2cc3c(c(=O)o2)C[C@]2(O)[C@@]4(C)CC[C@@H](O)C(C)(C)[C@]4(O)CC[C@@]2(C)O3)cc1', 'O=C(CCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2)NC12CC3CC(F)(CC(C1)c1ccccc13)C2', 'Cc1c(C)c2ccc(OCC3CCN(CCCO)CC3)cc2oc1=O', 'Cc1c(C)c2ccc(OCC3CCN(CCCO)CC3)cc2oc1=O', 'CCOc1ccc(Cc2nc3cc(C(=O)NCCCCNc4c5c(nc6ccccc46)CCCC5)ccc3n2CCC(C)C)cc1', 'O=C1N/C(=N\\c2ccc3oc(=O)ccc3c2)S/C1=C/c1ccc(Cl)cc1', 'COc1ccc(C2=NN(C(C)=O)C(c3ccc(N4CCN(Cc5ccccc5)CC4)cc3)C2)cc1OC', 'COc1ccc(/C=C2/S/C(=N/c3ccc4oc(=O)ccc4c3)NC2=O)cc1', 'Cl.Cl.N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'COc1cc2oc(-c3ccc(NCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)cc3)cc(=O)c2c(OC)c1OC', 'C(=N/Nc1c2c(nc3ccccc13)CCCC2)\\c1ccc(OCc2cn(Cc3ccccc3)nn2)cc1.Cl', 'COC(=O)/C(=C\\C(=O)c1cc(O)ccc1O)CC/C=C(\\C)CCC=C(C)C', 'COc1cc(NCc2ccc(Oc3c(N)cccc3Cl)cc2)cc(OC)c1', 'O=C(c1c(F)ccc(F)c1F)n1nc(OCCN2CCCCC2)c2ccccc21', 'O=C(CNc1ccc2oc(=O)ccc2c1)N1CCN(c2ccccc2)CC1', 'COc1ccc2nc3c(c(NCCCCCCNC(=O)[C@@H](N)Cc4c[nH]c5ccccc45)c2c1)CCCC3.Cl.Cl', 'Cc1ccc2[nH]c3c(c2c1)CN(CCCCCN1CCc2[nH]c4ccc(C)cc4c2C1)CC3', 'Oc1ccccc1/C=N/C1CCN(Cc2ccccc2)CC1', 'COc1ccc2nc3c(c(NCCCCCNC(=S)NC45CC6CC(CC(C6)C4)C5)c2c1)CCCC3', 'FC12CC3CC(NCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)(CC(C1)c1ccccc13)C2', 'Cc1c(C)c2ccc(OCC3CCN(Cc4cccc(OCCCO)c4)CC3)cc2oc1=O', 'NC12CC3CC(F)(CC(C1)c1cc(NC(=O)CCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)ccc13)C2', 'COc1ccc2nc3c(c(NCCCCCNC(=O)[C@@H](N)Cc4c[nH]c5ccccc45)c2c1)CCCC3.Cl.Cl', 'Fc1ccc2[nH]c3c(c2c1)C[PH](Cc1cccc2ccccc12)(c1ccccc1)CC3', 'Cc1cc(C)c(C(=O)n2nc(OCCN3CCCCC3)c3ccccc32)c(C)c1', 'C#CCNC1CCc2cc(Cl)c(OC(=O)N(C)CCC)cc21', 'CCOc1ccc(Cc2nc3cc(C(=O)N(CC)CC)ccc3n2CCCCNc2c3c(nc4ccccc24)CCCC3)cc1', 'O=C(CNc1ccc2oc(=O)ccc2c1)N1CCOCC1', 'Cc1ccc2oc(-c3ccc(OCCCCCCN4CCCCC4)cc3)cc(=O)c2c1', 'Cc1ccc2oc(-c3ccc(OCCCCCCN4CCCCC4)cc3)cc(=O)c2c1', 'Cc1ccc2oc(-c3ccc(OCCCCCCN4CCCCC4)cc3)cc(=O)c2c1', 'COc1ccc(CCOCCCCCNc2c3c(nc4ccccc24)CCCC3)cc1', 'COc1ccc2nc3c(c(NCCCCNC(=O)[C@@H](N)Cc4c[nH]c5ccccc45)c2c1)CCCC3.Cl.Cl', 'O=C(/C=C/c1ccccc1)c1ccc(OCCCCCCN2CCCC2)cc1O', 'O=C(NNC(=S)Nc1ccc(Cl)c(Cl)c1)c1ccccc1O', 'COc1ccc(-c2nc(NC(=O)C3CCN(Cc4cc(F)cc(F)c4)CC3)sc2-c2ccc(OC)cc2)cc1', 'O=C(Cc1cc(=O)oc2cc(O)ccc12)NCCCNCCNCCCNc1c2c(nc3ccccc13)CCCC2', 'Nc1cccc(Cl)c1Oc1ccc(CNc2ccc(O)cc2)cc1', 'O=C(CNc1ccc(F)cc1)Nc1ccc2oc(=O)ccc2c1', 'O=c1ccc2cc(/N=C/c3ccc(Br)cc3)ccc2o1', 'O=C(Nc1cc(COCc2cn(CCNc3c4c(nc5ccccc35)CCCC4)nn2)ccn1)Nc1cccc2c1C1CCCCN1C2=O', 'Fc1ccc(CN2CCC(Nc3nc(-c4ccc(Cl)cc4)c(-c4ccc(Cl)cc4)s3)CC2)c(Cl)c1', 'Fc1ccc(/C=N/C2CCN(Cc3ccccc3)CC2)cc1', 'Cc1ccc(-c2nc(NC3CCN(Cc4ccccc4)CC3)sc2-c2ccc(C)cc2)cc1', 'O=C1N/C(=N\\c2ccc3oc(=O)ccc3c2)S/C1=C/c1ccc(Br)cc1', 'CSc1ccc(C2SCC(=O)N2c2ccc3oc(=O)ccc3c2)cc1', 'Brc1ccc(-c2nnn3c2C2c4ccccc4CCN2Cc2ccccc2-3)cc1', 'COc1ccc(-c2nc(NC(=O)C3CCN(Cc4ccc(F)cc4)CC3)sc2-c2ccc(OC)cc2)cc1', 'COc1ccc(/C=N/c2ccc3oc(=O)ccc3c2)cc1', 'Oc1cc(/C=N/c2c3c(nc4ccccc24)CCCC3)cc(O)c1O', 'O=C(/C=C/c1ccc(OCc2ccccc2)cc1)NCCNc1c2c(nc3ccccc13)CCCC2', 'CC1=C[C@@H]2Cc3nc4cc(F)ccc4c(N)c3[C@H](C1)C2', 'NS(=O)(=O)c1ccc(C(=O)n2[nH]c(=O)c3c([N+](=O)[O-])cccc3c2=O)cc1', 'COc1ccc(NC(=S)NNc2c3c(nc4ccccc24)CCCC3)cc1', 'COc1cc(N)c(Cl)cc1C(=O)CCC1CCN(Cc2ccc(O)c(O)c2)CC1', 'COc1c(C(=O)CCC2CCN(Cc3ccccc3)CC2)cc(Cl)c2[nH]ccc12', 'Brc1ccc(-n2cc(COc3ccc(/C=N/Nc4c5c(nc6ccccc46)CCCC5)cc3)nn2)cc1.Cl', 'Clc1ccc(/C=N/C2CCN(Cc3ccccc3)CC2)cc1', 'COc1ccc(-c2nc(NC(=O)C3CCN(Cc4ccccc4C(F)(F)F)CC3)sc2-c2ccc(OC)cc2)cc1', 'COc1cc(CN2CCC(CCC(=O)c3cc(Cl)c(N)cc3OC)CC2)ccc1O', 'COc1ccc(-c2nc(NCC3CCN(Cc4ccccc4C(F)(F)F)CC3)sc2-c2ccc(OC)cc2)cc1', 'O=C(CCC1CCN(CC2CCCCC2)CC1)c1ccc2[nH]ccc2c1', 'Fc1cc(F)cc(CN2CCC(CNc3nc(-c4ccc(Cl)cc4)c(-c4ccc(Cl)cc4)s3)CC2)c1', 'O=C(Nc1nc(-c2ccc(Cl)cc2)c(-c2ccc(Cl)cc2)s1)C1CCN(Cc2ccc(F)cc2Cl)CC1', 'Cc1ccccc1CN1CCC(CNc2nc(-c3ccc(Cl)cc3)c(-c3ccc(Cl)cc3)s2)CC1', 'Cc1c(C)c2ccc(OCC3CCN(CCCO[N+](=O)[O-])CC3)cc2oc1=O', 'Oc1ccc(CCOCCCCCNc2c3c(nc4ccccc24)CCCC3)cc1', 'O=C(Nc1nc(-c2ccc(Cl)cc2)c(-c2ccc(Cl)cc2)s1)C1CCN(Cc2cc(F)cc(F)c2)CC1', 'Cc1ccc(-c2nc(NCC3CCN(Cc4ccccc4C(F)(F)F)CC3)sc2-c2ccc(C)cc2)cc1', 'COc1ccc(-c2cc(=O)c3cc(OC)c(O)c([C@@H]4CCC(=O)N4)c3o2)cc1', 'Cc1ccc(-c2nc(NCC3CCN(Cc4cc(F)cc(F)c4)CC3)sc2-c2ccc(C)cc2)cc1', 'O=C(/C=C/c1ccc([N+](=O)[O-])cc1)NCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'Fc1ccc(CN2CCC(CNc3nc(-c4ccc(Cl)cc4)c(-c4ccc(Cl)cc4)s3)CC2)c(C(F)(F)F)c1', 'Cc1ccc(-c2nc(NC(=O)C3CCN(Cc4ccccc4C(F)(F)F)CC3)sc2-c2ccc(C)cc2)cc1', 'COc1cc2ncnc(NCc3cn(CC4CCN(Cc5ccccc5)CC4)nn3)c2cc1OC', 'COc1cc(/C=C/C(=O)NCCNc2c3c(nc4ccccc24)CCCC3)ccc1OCc1ccccc1', 'Cc1ccc(-c2nc(NC(=O)C3CCN(Cc4cc(F)cc(F)c4)CC3)sc2-c2ccc(C)cc2)cc1', 'Oc1ccc(CNc2c3c(nc4ccccc24)CCCC3)cc1O', 'Cc1ccc(-c2nc(NC(=O)C3CCN(Cc4ccc(C#N)cc4)CC3)sc2-c2ccc(C)cc2)cc1', 'Oc1ccc(/C=N/c2c3c(nc4ccccc24)CCCC3)cc1O', 'COc1ccc2[nH]cc(CCNC(=O)CN(CCC3CCN(Cc4ccccc4)CC3)C(=O)/C=C/c3ccc(O)c(OC)c3)c2c1', 'O=C(Nc1cccc(-c2cn(CCNc3c4c(nc5ccccc35)CCCC4)nn2)n1)Nc1cccc2c1C1CCCCN1C2=O', 'COc1c(C(=O)CCC2CCN(CC3CCCCC3)CC2)cc(Cl)c2[nH]ccc12', 'O=C(/C=C/c1ccc(OCCCCCCN2CCCC2)cc1)c1ccc(OCCCCCCN2CCCC2)cc1O', 'C(=N/C1CCN(Cc2ccccc2)CC1)\\c1ccccc1', 'O=c1ccc2cc(/N=C/c3ccccc3)ccc2o1', 'COc1ccc2nc3c(c(NCCNC(=O)[C@@H](N)Cc4c[nH]c5ccccc45)c2c1)CCCC3.Cl.Cl', 'Cl.c1ccc(CCCCN2CCCc3[nH]c4ccccc4c3C2)cc1', 'O=C(c1ccccc1Cl)n1nc(OCCCN2CCCCC2)c2ccccc21', 'COc1c(C(=O)CCC2CCN(Cc3ccccc3)CC2)cc(Cl)c2[nH]c([Si](C)(C)C)cc12', 'Cl.Cl.N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'CN(C)c1ccc(/C=C2/S/C(=N/c3ccc4oc(=O)ccc4c3)NC2=O)cc1', 'Cc1ccc(CN2CCC(Nc3nc(-c4ccc(C)cc4)c(-c4ccc(C)cc4)s3)CC2)cc1', 'O=C(c1ccccc1Cl)n1nc(OCCN2CCCC2)c2ccccc21', 'COc1ccc2[nH]cc(CCNC(=O)CN(CCCC3CCN(Cc4ccccc4)CC3)C(=O)/C=C/c3ccc(O)c(OC)c3)c2c1', 'COc1c(C(=O)CCC2CCN(CC3CCCCC3)CC2)cc(Cl)c2c1ccn2Cc1ccccc1', 'Cl.Cl.N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'NS(=O)(=O)c1ccc(C(=O)n2[nH]c(=O)c3ccc(C(=O)O)cc3c2=O)cc1', 'C#CCOc1ccc(/C=N/Nc2c3c(nc4ccccc24)CCCC3)cc1.Cl', 'O=C1N/C(=N\\c2ccc3oc(=O)ccc3c2)S/C1=C/c1ccc(F)cc1', 'COc1ccc(-c2nc(NC3CCN(Cc4ccc(C(F)(F)F)cc4)CC3)sc2-c2ccc(OC)cc2)cc1', 'COc1cc2c(cc1OC)-c1cc(OCCN3CCCCC3)nc(=O)n1CC2', 'CCOc1ccc(Cc2nc3cc(C(=O)N(CC)CCCCNc4c5c(nc6ccccc46)CCCC5)ccc3n2CCC(C)C)cc1', 'c1ccc(COc2ccc(CCOCCCCCNc3c4c(nc5ccccc35)CCCC4)cc2OCc2ccccc2)cc1', 'N#Cc1ccccc1CN1CCC(C(=O)Nc2nc(-c3ccc(Cl)cc3)c(-c3ccc(Cl)cc3)s2)CC1', 'COc1cc2c(cc1O)C(Cc1ccc(O)c(Oc3cc(CC4c5cc(O)c(OC)cc5CCN4C)ccc3OC)c1)N(C)CC2', 'O=C(Nc1nc(-c2ccc(Cl)cc2)c(-c2ccc(Cl)cc2)s1)C1CCN(Cc2ccc(OC(F)(F)F)cc2)CC1', 'Cc1ccc(-c2nc(NCC3CCN(Cc4ccc(F)cc4)CC3)sc2-c2ccc(C)cc2)cc1', 'COc1ccc(-c2nc(NC(=O)C3CCN(Cc4ccc(F)cc4C(F)(F)F)CC3)sc2-c2ccc(OC)cc2)cc1', 'COc1ccc(CN2CCC(Nc3nc(-c4ccc(Cl)cc4)c(-c4ccc(Cl)cc4)s3)CC2)cc1', 'Cl.Cl.N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'COc1ccc2nc3c(c(NCCCCCCCCN(CC(=O)Nc4ccc5ccccc5c4)C(=O)/C=C/c4ccc(O)c(OC)c4)c2c1)CCCC3', 'COc1ccc(-n2cc(COc3ccc(/C=N/Nc4c5c(nc6ccccc46)CCCC5)cc3)nn2)cc1.Cl', 'CN1c2ccc(OC(=O)NCCCCCCN3CCCCC3)cc2CN2CCc3ccccc3C21', 'CN1c2ccc(OC(=O)NCCCCCCN3CCCCC3)cc2CN2CCc3ccccc3C21', 'Fc1ccc(CN2CCC(CNc3nc(-c4ccc(Cl)cc4)c(-c4ccc(Cl)cc4)s3)CC2)c(Cl)c1', 'COc1ccc(CCOCCCCCNc2c3c(nc4ccccc24)CCCC3)cc1OC', 'COc1ccc(-c2nc(NC3CCN(Cc4ccc(C)cc4)CC3)sc2-c2ccc(OC)cc2)cc1', 'Cl.Cl.N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'CCOc1ccc(Cc2nc3cc(C(=O)NCCOCCOCCNc4c5c(nc6ccccc46)CCCC5)ccc3n2CCC(C)C)cc1', 'Nc1cccc(Cl)c1Oc1ccc(CNc2cccc(O)c2)cc1', 'Cl.Cl.N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'Cl.Cl.N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'COc1ccc2nc3c(c(NCCCNC(=O)[C@@H](N)Cc4c[nH]c5ccccc45)c2c1)CCCC3.Cl.Cl', 'Cl.Cl.N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCCCCNc1c2c(nc3ccccc13)CCCC2', 'C=CCNC(=S)NNc1c2c(nc3ccccc13)CCCC2', 'COc1cc(O)c(C(C)=O)c(-c2c(C)cc(O)c3oc4c(O)cc(C)c(-c5cc(OC)cc(O)c5C(=O)O)c4c23)c1', 'Oc1ccc(/C=N/C2CCN(Cc3ccccc3)CC2)c(O)c1', 'Cl.O=C(O)c1ccc(-n2cc(COc3ccc(/C=N/Nc4c5c(nc6ccccc46)CCCC5)cc3)nn2)cc1', 'CCOc1ccc2[nH]c(SCc3cn(CC4CCN(Cc5ccccc5)CC4)nn3)nc2c1', 'CCOc1ccc(Cc2nc3cc(C(=O)NCCCCCCNc4c5c(nc6ccccc46)CCCC5)ccc3n2CCC(C)C)cc1', 'FC12CC3CC(NCCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)(CC(C1)c1ccccc13)C2', 'Cl.Cl.N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'Cl.Cl.NC(Cc1c[nH]c2ccccc12)C(=O)NCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'COc1ccc(-c2nc(NCC3CCN(Cc4c(F)cccc4Cl)CC3)sc2-c2ccc(OC)cc2)cc1', 'COc1c(C(=O)CCC2CCN(CC3CCCCC3)CC2)cc(Cl)c2c1ccn2C', 'O=C(Nc1cc(-c2cn(CCNc3c4c(nc5ccccc35)CCCC4)nn2)ccn1)Nc1cccc2c1C1CCCCN1C2=O', 'Cc1ccc(-c2nc(NCC3CCN(Cc4ccccc4C)CC3)sc2-c2ccc(C)cc2)cc1', 'C/C(=C/CC[C@H](CO)c1cc2cc(O)c(O)cc2o1)CO', 'Oc1ccc(CNC2CCN(Cc3ccccc3)CC2)cc1', 'O=C(CCC1CCN(CC2CCCC2)CC1)c1ccc2[nH]ccc2c1', 'COc1ccc(CN2CCC(Nc3nc(-c4ccc(OC)cc4)c(-c4ccc(OC)cc4)s3)CC2)cc1', 'Cl.Cl.N[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=C(c1cccc(F)c1F)n1nc(OCCN2CCCC2)c2ccccc21', 'COc1ccc(C2SCC(=O)N2c2ccc3oc(=O)ccc3c2)cc1', 'Cc1ccc(/C=N/C2CCN(Cc3ccccc3)CC2)cc1', 'Cc1ccccc1CN1CCC(C(=O)Nc2nc(-c3ccc(Cl)cc3)c(-c3ccc(Cl)cc3)s2)CC1', 'COc1ccc(-c2nc(NCC3CCN(Cc4ccccc4C)CC3)sc2-c2ccc(OC)cc2)cc1', 'COc1ccc2nc3c(c(NCCCCCCCCNC(=S)NC45CC6CC(CC(C6)C4)C5)c2c1)CCCC3', 'CCCCN1CCC(CCC(=O)c2ccc3[nH]ccc3c2)CC1', 'COc1cccc(NCc2ccc(Oc3c(N)cccc3Cl)cc2)c1', 'Clc1ccc(/C=N/C2CCN(Cc3ccccc3)CC2)c(Cl)c1', 'Fc1ccc(/C=N/C2CCN(Cc3ccccc3)CC2)c(F)c1', 'CC[C@@H](CN1CCCC1=O)C(=O)NCCCCCCCNc1c2c(nc3cc(Cl)ccc13)C[C@H]1C=C(C)C[C@@H]2C1', 'COc1cc2oc(-c3ccc(NCCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)cc3)cc(=O)c2c(OC)c1OC', 'COc1ccc(-c2nc(NC(=O)C3CCN(Cc4c(F)cccc4Cl)CC3)sc2-c2ccc(OC)cc2)cc1', 'O=c1ccc2cc(/N=C/c3ccc(F)cc3)ccc2o1', 'COc1cc(O)c2c(c1)C1=C(c3c(O)c(O)cc(C)c3-c3cc(OC)cc(O)c3C(=O)O)C(=O)C(O)=CC1(C)OC2=O', 'COc1ccc(-c2nc(NC(=O)C3CCN(Cc4cccc(F)c4)CC3)sc2-c2ccc(OC)cc2)cc1', 'Cc1c(C)c2ccc(OCC3CCCN(Cc4ccccc4)C3)cc2oc1=O', 'Cc1c(C)c2ccc(OCC3CCCN(Cc4ccccc4)C3)cc2oc1=O', 'Cc1c(C)c2ccc(OCC3CCCN(Cc4ccccc4)C3)cc2oc1=O', 'Cc1ccc(-c2nc(NC(=O)C3CCN(Cc4ccc([N+](=O)[O-])cc4)CC3)sc2-c2ccc(C)cc2)cc1', 'Oc1ccc(CNCCCCCCNc2c3c(nc4ccccc24)CCCC3)cc1O', 'Cc1ccc(-c2nc(NC(=O)C3CCN(Cc4c(F)cccc4Cl)CC3)sc2-c2ccc(C)cc2)cc1', 'O=C1c2ccccc2C(=O)N1Cc1cn(CC2CCN(Cc3ccccc3)CC2)nn1', 'FC(F)(F)c1ccc(/C=N/C2CCN(Cc3ccccc3)CC2)cc1', 'Clc1ccc(-c2nc(NC3CCN(Cc4ccccc4)CC3)sc2-c2ccc(Cl)cc2)cc1', 'COc1ccc(C2=NN(C(C)=O)C(c3ccc(N4CCC(Cc5ccccc5)CC4)cc3)C2)cc1OC', 'COc1ccc(C2CC(=O)c3ccc4c(c3O2)C=CC(C)(C)O4)cc1', 'Cc1c(C)c2ccc(OCC3CCN(Cc4cccc(OCCCO[N+](=O)[O-])c4)CC3)cc2oc1=O', 'O=C(CCC1CCN(CC2CCCCC2)CC1)c1ccc2c(ccn2Cc2ccccc2)c1', 'NS(=O)(=O)c1ccc(C(=O)n2[nH]c(=O)c3c(F)cccc3c2=O)cc1', 'Oc1ccc2[nH]c3c(c2c1)C[PH](Cc1cccc2ccccc12)(c1ccccc1)CC3', 'C#CCNC1CCCc2ccc(OC(=O)N(C)CC)cc21', 'S=C(NNc1c2c(nc3ccccc13)CCCC2)Nc1ccccc1', 'COc1ccc2nc3c(c(NCCCCCCCCNC(=O)[C@@H](N)Cc4c[nH]c5ccccc45)c2c1)CCCC3.Cl.Cl', 'COc1ccc(C2=NN(C(C)=O)C(c3ccc(N4CCN(C)CC4)cc3)C2)cc1OC', 'COc1ccc(C2=NN(C(C)=O)C(c3ccc(N4CCN(C)CC4)cc3)C2)cc1OC', 'CN1c2ccc(OC(=O)NCCCCCCN3CCCC3)cc2CN2CCc3ccccc3C21', 'CN1c2ccc(OC(=O)NCCCCCCN3CCCC3)cc2CN2CCc3ccccc3C21', 'FC(F)(F)c1ccc(CNC2CCN(Cc3ccccc3)CC2)cc1', 'FC(F)(F)c1ccc(CNC2CCN(Cc3ccccc3)CC2)cc1', 'O=C(Nc1ccc(-c2cn(CCNc3c4c(nc5ccccc35)CCCC4)nn2)cn1)Nc1cccc2c1C1CCCCN1C2=O', 'CCOc1ccc(Cc2nc3cc(C(=O)N(CC)CC)ccc3n2CCCCCNc2c3c(nc4ccccc24)CCCC3)cc1', 'COc1c(C(=O)CCC2CCN(CC3CCCCC3)CC2)cc(Cl)c2c1ccn2S(=O)(=O)c1ccccc1.O=C(O)/C=C/C(=O)O', 'O=C(CCC1CCN(Cc2ccccc2)CC1)c1ccc2[nH]ccc2c1', 'COc1ccc2nc3c(c(NCCCCCCCNC(=O)[C@@H](N)Cc4c[nH]c5ccccc45)c2c1)CCCC3.Cl.Cl', 'COc1ccc(-c2nc(NC(=O)C3CCN(Cc4ccccc4C#N)CC3)sc2-c2ccc(OC)cc2)cc1', 'FC(F)(F)Oc1ccc(CNC2CCN(Cc3ccccc3)CC2)cc1', 'Cc1ccc(CN2CCC(Nc3nc(-c4ccc(Cl)cc4)c(-c4ccc(Cl)cc4)s3)CC2)cc1', 'c1ccc(COc2ccc(CCOCCCCCNc3c4c(nc5ccccc35)CCCC4)cc2)cc1', 'CCOc1ccc(Cc2nc3cc(C(=O)NCCNc4c5c(nc6ccccc46)CCCC5)ccc3n2CCC(C)C)cc1', 'O=C(c1cccc(Cl)c1Cl)n1nc(OCCCN2CCCCC2)c2ccccc21', 'Cc1c(OCc2ccccc2)cccc1N1CCN(CCCCCCNc2c3c(nc4c2CCCC4)CCCC3)CC1', 'Oc1ccc(CCOCCCCCNc2c3c(nc4ccccc24)CCCC3)cc1O', 'CCN(CC)Cc1cc2cc(c1O)Oc1ccc(cc1)C[C@H]1c3cc(c(OC)cc3CCN1C)Oc1c(OC)c(OC)cc3c1[C@@H](C2)N(C)CC3', 'NC12CC3CC(F)(CC(C1)c1cc(NC(=O)CCCCNc4c5c(nc6cc(Cl)ccc46)CCCC5)ccc13)C2', 'COc1cc(CNC2CCN(Cc3ccccc3)CC2)cc(OC)c1OC', 'Cc1cc(Nc2nccc(-c3sc(C4CCNCC4)nc3-c3ccc(F)cc3)n2)ccn1', 'COc1ccccc1CN1CCN(CCCCCOc2cccc3[nH]c4ccccc4c23)CC1', 'O=C(Nc1nc(-c2ccc(Cl)cc2)c(-c2ccc(Cl)cc2)s1)C1CCN(Cc2cccc(F)c2)CC1', 'Cc1ccc(-n2cc(COc3ccc(/C=N/Nc4c5c(nc6ccccc46)CCCC5)cc3)nn2)cc1.Cl', 'O=C(Nc1nc(-c2ccc(Cl)cc2)c(-c2ccc(Cl)cc2)s1)C1CCN(Cc2ccccc2C(F)(F)F)CC1', 'O=C(Nc1cc(CCc2cn(CCNc3c4c(nc5ccccc35)CCCC4)nn2)ccn1)Nc1cccc2c1C1CCCCN1C2=O', 'COc1ccc(-c2nc(NC(=O)C3CCN(Cc4ccccc4C)CC3)sc2-c2ccc(OC)cc2)cc1', 'N#Cc1ccc(CN2CCC(Nc3nc(-c4ccc(Cl)cc4)c(-c4ccc(Cl)cc4)s3)CC2)cc1', 'CCN1CCN(c2ccc(C3CC(c4ccc(OC)c(OC)c4)=NN3C(C)=O)cc2)CC1', 'Cc1c(C)c2ccc(OCC3CCN(Cc4ccccc4)CC3)cc2oc1=O', 'COc1ccc([C@H]2OC[C@H]3[C@@H]2C(=O)O[C@@H]3c2ccc3c(c2)OCO3)cc1OC', 'O=C1CN(/N=C/c2ccc(-c3cccc([N+](=O)[O-])c3)o2)C(=O)N1CCC1CCN(Cc2ccccc2)CC1', 'O=[N+]([O-])C1C(Cc2c[nH]cn2)NC2(c3ccccc3-c3nc4ccccc4nc32)C1c1ccccc1', 'C#CCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'O=[N+]([O-])C1C(Cc2c[nH]cn2)NC2(c3ccccc3-c3nc4ccccc4nc32)C1c1ccco1', 'COc1cccc(-c2ccc(/C=N/N3CC(=O)N(CCC4CCN(Cc5ccccc5)CC4)C3=O)o2)c1', 'COc1ccc([C@H]2O[C@@H](O)[C@H]3[C@@H]2C(=O)O[C@@H]3c2ccc(OC)c(OC)c2)cc1OC', 'C#CCNc1c2c(nc3ccccc13)CCCC2', 'COc1ccc([C@H]2OC(=O)[C@H]3[C@@H]2C(=O)O[C@@H]3c2ccc3c(c2)OCO3)cc1OC', 'Cc1cccc(C2C([N+](=O)[O-])C(Cc3c[nH]cn3)NC23c2ccccc2-c2nc4ccccc4nc23)c1', 'COc1c2c(cc3c1OCO3)[C@]13C=C[C@H](OC)C[C@@H]1[N@](C2)C[C@H]3OC(=O)c1ccccc1[N+](=O)[O-]', 'COc1ccc([C@H]2O[C@@H](O)[C@H]3[C@@H]2C(=O)O[C@@H]3c2cc(OC)c3c(c2)OCO3)cc1OC', 'O=C1CN(/N=C/c2ccc(-c3ccc([N+](=O)[O-])cc3)o2)C(=O)N1CCC1CCN(Cc2ccccc2)CC1', 'O=[N+]([O-])C1C(Cc2c[nH]cn2)NC2(c3ccccc3-c3nc4ccccc4nc32)C1c1ccccn1', 'CC(C)=CC[C@H]1C[C@]2(CC=C(C)C)C(=O)[C@]3(CC=C(C)C)O[C@@]1(C(=O)C(C)C)[C@@]2(O)C(C)(C)C3=O', 'COc1cc2cc(NC(=O)CCCNCc3ccccc3)c(=O)oc2cc1OC', 'COc1ccc(-c2ccc(/C=N/N3CC(=O)N(CCC4CCN(Cc5ccccc5)CC4)C3=O)o2)cc1', 'COc1cccc(C2C([N+](=O)[O-])C(Cc3c[nH]cn3)NC23c2ccccc2-c2nc4ccccc4nc23)c1', 'Cc1ccc(C2C([N+](=O)[O-])C(Cc3c[nH]cn3)NC23c2ccccc2-c2nc4ccccc4nc23)cc1', 'O=C1CN(/N=C/c2ccc(-c3ccc([N+](=O)[O-])cc3)o2)C(=O)N1CCN1CCN(Cc2ccccc2)CC1', 'COc1c2c(cc3c1OCO3)[C@]13C=C[C@H](OC)C[C@@H]1[N@](C2)C[C@H]3OC(=O)c1ccccc1C', 'COc1ccccc1-c1ccc(/C=N/N2CC(=O)N(CCC3CCN(Cc4ccccc4)CC3)C2=O)o1', 'COc1ccc(C2C([N+](=O)[O-])C(Cc3c[nH]cn3)NC23c2ccccc2-c2nc4ccccc4nc23)cc1', 'COc1ccccc1C1C([N+](=O)[O-])C(Cc2c[nH]cn2)NC12c1ccccc1-c1nc3ccccc3nc12', 'O=C1CN(/N=C/c2ccc(-c3ccccc3[N+](=O)[O-])o2)C(=O)N1CCC1CCN(Cc2ccccc2)CC1', 'Cc1ccccc1C1C([N+](=O)[O-])C(Cc2c[nH]cn2)NC12c1ccccc1-c1nc3ccccc3nc12', 'COc1ccc(CCC(=O)Nc2nc(-c3cc4ccccc4oc3=O)cs2)cc1OC', 'COc1cc2cc(NC(=O)C3CCN(Cc4ccccc4)CC3)c(=O)oc2cc1OC', 'O=[N+]([O-])C1C(Cc2c[nH]cn2)NC2(c3ccccc3-c3nc4ccccc4nc32)C1c1ccc(Cl)cc1', 'COc1c2c(cc3c1OCO3)[C@]13C=C[C@H](OC)C[C@@H]1[N@](C2)C[C@H]3OC(=O)c1cccc([N+](=O)[O-])c1', 'COc1ccc([C@H](O)[C@H](N)C(=O)O)cc1OC', 'O=[N+]([O-])C1C(Cc2c[nH]cn2)NC2(c3ccccc3-c3nc4ccccc4nc32)C1c1ccc(Br)cc1', 'COc1cc2cc(NC(=O)[C@@H]3CCC[C@H](NCc4ccccc4)C3)c(=O)oc2cc1OC', 'COc1cc2cc(NC(=O)[C@@H]3CCC[C@H](NCc4ccccc4)C3)c(=O)oc2cc1OC', 'CN1C[C@@H](c2cccc(Br)c2)[C@@]2(CSc3ccccc3C2=O)C12C(=O)c1ccccc1C2=O', 'CCN(CC)C(=O)Oc1ccc(/C=C/C(=O)N2CCN(Cc3ccccc3)CC2)cc1OC', 'O=C(N/N=C/c1ccc(Cl)cc1)N1CCN(c2ccccn2)CC1', 'CN1C[C@@H](c2ccc(Br)cc2)[C@@]2(CSc3ccccc3C2=O)C12C(=O)c1ccccc1C2=O', 'COc1cc(/C=C/C(=O)N2CCc3ccccc3C2)ccc1OC(=O)N(C)OC', 'Cc1cc(S(=O)(=O)Nc2cnccn2)c(C(C)C)cc1O', 'COc1cc(CNC(=O)CCCC/C=C/CNc2c3c(nc4cc(Cl)ccc24)CC2C=C(C)CC3C2)ccc1O', 'CNC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCc1cn(CCNc2c3c(nc4cc(Cl)ccc24)CCCC3)nn1', 'Brc1ccc(CNC2CCN(Cc3ccccc3)C2)cc1', 'CNC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCC1CCN(Cc2ccc(F)c(F)c2)CC1', 'NC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCC1CCN(Cc2ccccc2Cl)CC1', 'O=C(CCCN1CCC2c3ccccc3NC2C1)Oc1ccc(/C=C2\\CCn3c2nc2ccccc2c3=O)cc1', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(NC(C)=O)cc3)CC2)nc1-c1ccc(OC)cc1', 'O=[N+]([O-])c1ccccc1-c1nnc(Nc2ncccn2)o1', 'COc1cc(CNC(=O)CCCCNc2c3c(nc4cc(Cl)ccc24)CC2C=C(C)CC3C2)ccc1O', 'Cc1ccc(-c2nnc(N3CCN(c4ccccn4)CC3)o2)cc1', 'COc1ccc2cc3[n+](cc2c1OCc1cccc(Br)c1)CCc1cc2c(cc1-3)OCO2.[Br-]', 'O=C(N/N=C/c1ccccc1)N1CCN(c2ccccn2)CC1', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(N)c(Cl)c3)CC2)nc1-c1ccc(OC)cc1', 'O=C1c2ccccc2C(=O)C12N1CSC[C@@H]1[C@@H](c1ccc(Cl)cc1)[C@@]21CSc2ccccc2C1=O', 'COc1cccc(-c2nnc(N3CCN(c4ccccn4)CC3)o2)c1OC', 'O=C(N/N=C/c1ccc([N+](=O)[O-])cc1)N1CCN(c2ccccn2)CC1', 'CN1C[C@@H](c2ccc(Cl)cc2)[C@@]2(CSc3ccccc3C2=O)C12C(=O)c1ccccc1C2=O', 'Fc1ccc(/C=N\\C2CCN(Cc3ccccc3)C2)c(F)c1', 'CCN(C)C(=O)Oc1cccc(CNCCc2c[nH]c3cc(F)ccc23)c1', 'COc1cc(CNC(=O)Cc2cn(CCCCC3=CC4Cc5nc6cc(Cl)ccc6c(N)c5C(C3)C4)nn2)ccc1O', 'COc1cc(/C=C/C(=O)N2CCc3ccccc3C2)ccc1OC(=O)N(c1ccccc1)c1ccccc1', 'COc1ccc(-c2nnc(Nc3ncccn3)o2)cc1', 'COc1ccc2cc3[n+](cc2c1OCc1ccc([N+](=O)[O-])cc1)CCc1cc2c(cc1-3)OCO2.[Br-]', 'COc1cc(/C=N\\C2CCN(Cc3ccccc3)C2)cc(OC)c1OC', 'COc1cc(-c2nnc(N3CCN(c4ccccn4)CC3)o2)cc(OC)c1OC', 'CCCCN[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCC1CCN(CC2CCN(C)CC2)CC1', 'COc1ccc2cc3[n+](cc2c1OCc1ccccc1Br)CCc1cc2c(cc1-3)OCO2.[Br-]', 'COc1ccc2cc3[n+](cc2c1OCc1ccc(C)cc1)CCc1cc2c(cc1-3)OCO2.[Br-]', 'Oc1ccc(-c2nnc(N3CCN(c4ccccn4)CC3)o2)cc1', 'CS[C@@H](Cn1nnc2cc(C#Cc3ccc(CN4CCOCC4)cc3)ccc21)N(O)C=O', 'CCCNC(=O)c1sc(-c2ccnc(NC(=O)CNCCNc3c4c(nc5ccccc35)CCCC4)c2)nc1OCC1CC1', 'CNC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCC1CCN(Cc2ccc(Cl)c(Cl)c2)CC1', 'c1ccc(-c2nnc(N3CCN(c4ccccn4)CC3)o2)cc1', 'O=C(N/N=C/c1ccc(Br)cc1)N1CCN(c2ccccn2)CC1', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(N)c(OC)c3)CC2)nc1-c1ccc(OC)c(OC)c1', 'COc1cc2c(cc1OCc1ccccc1)CCC(C)(CC(=O)NCCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'Brc1cccc(-c2nnc(Nc3ncccn3)o2)c1', 'COc1ccc2cc3[n+](cc2c1OCc1ccc(Cl)c(Cl)c1)CCc1cc2c(cc1-3)OCO2.[Br-]', 'COc1ccc([C@@H]2CN(C)C3(C(=O)c4ccccc4C3=O)[C@]23CSc2ccccc2C3=O)cc1', 'COc1cccc(CNC2CCN(Cc3ccccc3)C2)c1OC', 'COc1ccc2cc3[n+](cc2c1OCc1ccc2ccccc2c1)CCc1cc2c(cc1-3)OCO2.[Br-]', 'CCN(C)C(=O)Oc1ccc(/C=C/C(=O)N2CCCCC2)cc1OC', 'COc1cc2c(cc1OCc1ccccc1)CCC(C)(CC(=O)NCCCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'NC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCCC1CCN(Cc2ccc(F)cc2)CC1', 'O=C(N/N=C/c1ccc(C(F)(F)F)cc1)N1CCN(c2ccccn2)CC1', 'COc1cc(CNC(=O)CCCCCCCNc2c3c(nc4cc(Cl)ccc24)CC2C=C(C)CC3C2)ccc1O', 'O=C1c2ccccc2C(=O)C12N1CSC[C@@H]1[C@@H](c1ccccc1Cl)[C@@]21CSc2ccccc2C1=O', 'CNC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCC1CCN(Cc2ccccc2)CC1', 'COC(=O)c1ccc(COc2c(OC)ccc3cc4[n+](cc23)CCc2cc3c(cc2-4)OCO3)cc1.[Br-]', 'COc1cc(/C=C/C(=O)N2CCCCC2)ccc1OC(=O)N(C)C', 'NC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCCC1CCN(Cc2ccccc2F)CC1', 'COc1cc2c(cc1O)CCC(C)(CC(=O)NCCCCCCNc1c3c(nc4ccccc14)CCCC3)O2', 'Cc1cc(/C=N/O)c(O)c(CN2CCN(c3ccncc3)CC2)c1', 'Cc1cc(/C=N/O)c(O)c(CN2CCN(c3ccncc3)CC2)c1', 'CNC(=O)c1sc(-c2ccnc(NC(=O)CNCCNc3c4c(nc5ccccc35)CCCC4)c2)nc1OC', 'COc1ccc2cc3[n+](cc2c1OCc1ccc(Cl)cc1)CCc1cc2c(cc1-3)OCO2.[Br-]', 'NC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCC1CCN(Cc2ccccn2)CC1', 'CNC(=O)c1sc(-c2ccnc(NC(=O)CNCCNc3c4c(nc5ccccc35)CCCC4)c2)nc1OCC1CCC1', 'Fc1ccc(CNC2CCN(Cc3ccccc3)C2)cc1', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(NC(C)=O)c(Cl)c3)CC2)nc1-c1ccccc1', 'COCCn1cnc2cc(NC(=O)c3ccccc3)cc(C(=O)NCc3cc(F)cc(F)c3)c21', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(N)cc3)CC2)nc1-c1ccc(OC)c(OC)c1', 'CNC(=O)c1sc(-c2ccnc(NC(=O)CNCCNc3c4c(nc5ccccc35)CCCC4)c2)nc1OCC(C)C', 'CCN(C)C(=O)Oc1ccc(/C=C/C(=O)N2CCN(Cc3ccccc3)CC2)cc1', 'O=C(CCCCCN1CCC2c3ccccc3NC2C1)Oc1ccc(/C=C2\\CCn3c2nc2ccccc2c3=O)cc1', 'CCCCN[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCC1CCCN(Cc2ccccc2)CC1', 'COc1ccc2nc3cc(Cl)ccc3c(Nc3ccc(C(=O)NO)cc3)c2n1', 'O=C(NO)c1ccc(Nc2c3ccccc3nc3ccc(OC(F)(F)F)cc23)cc1', 'COc1cc2c(cc1O)CCC(C)(CCNCCCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'COc1cc2c(cc1O)CCC(C)(CCNCCCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'COc1cc2c(cc1O)CCC(C)(CCNCCCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'C[C@@H]1CC[C@H]2[C@H](C(=O)c3c(O)c([C@@]4(O)CC[C@@H](O)[C@H]5O[C@H]54)cn(O)c3=O)[C@H](C)C=C[C@@H]2C1', 'COc1ccc2cc3[n+](cc2c1OCc1ccc(C(F)(F)F)cc1)CCc1cc2c(cc1-3)OCO2.[Br-]', 'COc1ccc2cc3[n+](cc2c1OCc1ccc(C(F)(F)F)cc1)CCc1cc2c(cc1-3)OCO2.[Br-]', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(NC(C)=O)c(F)c3)CC2)nc1-c1ccc(OC)c(OC)c1', 'CNC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OC1CCN(Cc2ccccc2)CC1', 'COc1ccc([C@@H]2[C@H]3CSCN3C3(C(=O)c4ccccc4C3=O)[C@]23CSc2ccccc2C3=O)cc1', 'Cc1cccc(-c2nnc(N3CCN(c4ccccn4)CC3)o2)c1', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(NC(C)=O)c(F)c3)CC2)nc1-c1ccccc1', 'O=C(NO)c1ccc(Nc2ccnc3ccccc23)cc1', 'CO[C@]1(c2cn(O)c(=O)c(C(=O)[C@H]3[C@@H]4CC[C@@H](C)C[C@H]4C=C[C@H]3C)c2O)CC[C@@H](O)[C@H]2O[C@H]21', 'Fc1ccc(CNC2CCN(Cc3ccccc3)C2)c(F)c1', 'COc1ccc2cc3[n+](cc2c1OC(C)c1ccccc1)CCc1cc2c(cc1-3)OCO2.[Br-]', 'Cc1cc(/C=N/O)c(O)c(CNc2ccc(Cn3cncn3)cc2)c1', 'Cc1cc(/C=N/O)c(O)c(CNc2ccc(Cn3cncn3)cc2)c1', 'Cc1cc(/C=N/O)c(O)c(CNc2ccc(Cn3cncn3)cc2)c1', 'Cc1cc(/C=N/O)c(O)c(CNc2ccc(Cn3cncn3)cc2)c1', 'CNC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCC1CCN(Cc2ccccc2F)CC1', 'NC(=O)c1sc(-c2ccnc(NC(=O)CNCCNc3c4c(nc5ccccc35)CCCC4)c2)nc1OCC1CC1', 'NC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCC1CCN(Cc2ccccc2)CC1', 'COc1ccc2cc3[n+](cc2c1OCc1cccc(Cl)c1)CCc1cc2c(cc1-3)OCO2.[Br-]', 'COc1cc2nc3ccccc3c(Nc3ccc(C(=O)NO)cc3)c2c(OC)c1OC', 'O=C1c2ccccc2C(=O)C12N1CSC[C@@H]1[C@@H](c1ccccc1)[C@@]21CSc2ccccc2C1=O', 'O=c1ccoc2ccc(OCCCN3CCCCCC3)cc12', 'O=c1ccoc2ccc(OCCCN3CCCCCC3)cc12', 'O=C(NCCCNc1c2c(nc3ccccc13)CCCC2)c1sc(-c2ccnc(NC(=O)C3CCC3)c2)nc1OCC1CC1', 'O=[N+]([O-])c1ccc(-c2nnc(N3CCN(c4ccccn4)CC3)o2)cc1', 'COc1cc(CNC(=O)CCCCCCNc2c3c(nc4cc(Cl)ccc24)CC2C=C(C)CC3C2)ccc1O', 'COc1cc(CNC(=O)CCCCCCNc2c3c(nc4cc(Cl)ccc24)CC2C=C(C)CC3C2)ccc1O', 'COc1cc(CNC(=O)CCCCCCNc2c3c(nc4cc(Cl)ccc24)CC2C=C(C)CC3C2)ccc1O', 'COc1cc(CNC(=O)CCCCCCNc2c3c(nc4cc(Cl)ccc24)CC2C=C(C)CC3C2)ccc1O', 'CN1C[C@@H](c2ccccc2Cl)[C@@]2(CSc3ccccc3C2=O)C12C(=O)c1ccccc1C2=O', 'Oc1ccc(/C=N\\C2CCN(Cc3ccccc3)C2)cc1', 'FC(F)(F)Oc1ccc(/C=N\\C2CCN(Cc3ccccc3)C2)cc1', 'NC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCC1CCN(Cc2c(F)cccc2F)CC1', 'Fc1ccc(-c2nnc(Nc3ncccn3)o2)cc1', 'CN1C[C@@H](c2ccccc2Br)[C@@]2(CSc3ccccc3C2=O)C12C(=O)c1ccccc1C2=O', 'O=C(CCN1CCC2c3ccccc3NC2C1)Oc1ccc(/C=C2\\CCn3c2nc2ccccc2c3=O)cc1', 'COc1cc(/C=C/C(=O)N2CCCCC2)ccc1OC(=O)N(c1ccccc1)c1ccccc1', 'COc1cc(/C=C/C(=O)N2CCN(Cc3ccccc3)CC2)ccc1OC(=O)N(C)C', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(NC(C)=O)c(OC)c3)CC2)nc1-c1ccc(OC)cc1', 'CNC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCOCc1cn(CCNc2c3c(nc4cc(Cl)ccc24)CCCC3)nn1', 'NC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCC1CCN(Cc2ccc(F)c(F)c2)CC1', 'COc1ccc2cc3[n+](cc2c1OCc1ccccc1C)CCc1cc2c(cc1-3)OCO2.[Br-]', 'Clc1ccc(-c2nnc(N3CCN(c4ccccn4)CC3)o2)cc1', 'CC(C)c1ccc([C@@H]2[C@H]3CSCN3C3(C(=O)c4ccccc4C3=O)[C@]23CSc2ccccc2C3=O)cc1', 'O=C(N/N=C/c1ccc(O)cc1O)N1CCN(c2ccccn2)CC1', 'Oc1ccc(CNC2CCN(Cc3ccccc3)C2)c(O)c1', 'CNC(=O)c1sc(-c2ccnc(NC(=O)CNCCCCNc3c4c(nc5ccccc35)CCCC4)c2)nc1OCC1CC1', 'O=C(NO)c1ccc(Nc2c3ccccc3nc3ccc(F)cc23)cc1', 'COc1cc(CNC2CCN(Cc3ccccc3)C2)cc(OC)c1OC', 'O=[N+]([O-])c1ccc(/C=N\\C2CCN(Cc3ccccc3)C2)cc1', 'O=C(CCN1CCC2c3ccccc3NC2C1)Nc1ccc(/C=C2\\CCn3c2nc2ccccc2c3=O)cc1', 'COc1cc2c(cc1O)CCC(C)(CC(=O)NCCCCCCCCCNc1c3c(nc4ccccc14)CCCC3)O2', 'O=C1c2ccccc2C(=O)C12N1CSC[C@@H]1[C@@H](c1cccc3ccccc13)[C@@]21CSc2ccccc2C1=O', 'COc1ccc2cc3[n+](cc2c1OCc1ccc(C(C)C)cc1)CCc1cc2c(cc1-3)OCO2.[Br-]', 'O/N=C/c1cc(Cl)cc(CN2CCN(c3ccncc3)CC2)c1O', 'O/N=C/c1cc(Cl)cc(CN2CCN(c3ccncc3)CC2)c1O', 'COc1ccc2cc3[n+](cc2c1OCc1cccc(C)c1)CCc1cc2c(cc1-3)OCO2.[Br-]', 'CC(=O)Nc1cc(-c2nc(OCC3CC3)c(C(=O)NCCCNc3c4c(nc5ccccc35)CCCC4)s2)ccn1', 'COc1cc2c(cc1O)CCC(C)(CC(=O)NCCOCCOCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'c1ccc(CNC2CCN(Cc3ccccc3)C2)cc1', 'CS[C@@H](Cn1nnc2cc(C#Cc3ccc(N4CC5(COC5)C4)cc3)ccc21)N(O)C=O', 'C/C=C1\\C2C=C(C)CC1(NCC(=O)OC)c1ccc(=O)[nH]c1C2', 'Fc1ccc(-c2nnc(N3CCN(c4ccccn4)CC3)o2)c(F)c1', 'Fc1ccc(-c2nnc(N3CCN(c4ccccn4)CC3)o2)c(F)c1', 'COc1ccc2cc3[n+](cc2c1OCc1cccc([N+](=O)[O-])c1)CCc1cc2c(cc1-3)OCO2.[Br-]', 'CNC(=O)c1sc(-c2ccnc(NC(=O)CNCCNc3c4c(nc5ccccc35)CCCC4)c2)nc1OCC(C)(C)C', 'NC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCC1CCN(Cc2ccccc2F)CC1', 'COc1cc(/C=C/C(=O)N2CCC(Cc3ccccc3)CC2)ccc1OC(=O)N(c1ccccc1)c1ccccc1', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(NC(C)=O)cc3)CC2)nc1-c1ccccc1', 'COc1ccc2[nH]cc(CCNC(=O)CN(CCCCCCCCNc3c4c(nc5ccccc35)CCCC4)C(=O)/C=C/c3ccc(O)c(OC)c3)c2c1', 'COc1cc(/C=C/C(=O)N2CCc3ccccc3C2)ccc1OC(=O)N(C)C', 'CNC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCCc1cn(CCNc2c3c(nc4ccccc24)CCCC3)nn1', 'Cc1ccc([C@@H]2[C@H]3CSCN3C3(C(=O)c4ccccc4C3=O)[C@]23CSc2ccccc2C3=O)cc1', 'CC1=C[C@@H]2Cc3[nH]c(=O)ccc3[C@]3(C1)NCCC[C@H]23', 'Clc1ccc(CNC2CCN(Cc3ccccc3)C2)cc1', 'c1cnc(Nc2nnc(-c3cccc4ccccc34)o2)nc1', 'c1cnc(Nc2nnc(-c3cccc4ccccc34)o2)nc1', 'CCN(CC)C(=O)Oc1ccc(/C=C/C(=O)N2CCC(Cc3ccccc3)CC2)cc1OC', 'CCN(C)C(=O)Oc1cccc(CNCCc2c[nH]c3cc(F)ccc23)c1.O=C(O)/C=C/C(=O)O', 'CCNC(=O)c1sc(-c2ccnc(NC(=O)CNCCNc3c4c(nc5ccccc35)CCCC4)c2)nc1OCC1CC1', 'FC(F)(F)c1ccc(CNC2CCN(Cc3ccccc3)C2)cc1', 'Oc1ccc(-c2nnc(N3CCN(c4ccccn4)CC3)o2)c(O)c1', 'O=C1c2ccccc2C(=O)C12N1CSC[C@@H]1[C@@H](c1cccc(F)c1)[C@@]21CSc2ccccc2C1=O', 'COc1ccc2cc3[n+](cc2c1OCc1ccc(Br)cc1)CCc1cc2c(cc1-3)OCO2.[Br-]', 'O=C(NO)c1ccc(Nc2c3ccccc3nc3ccc(OCc4ccccc4)cc23)cc1', 'CCN(CC)c1cc(/C=C2\\OC(=O)c3cc(OC)c(OC)cc32)ccc1O', 'Fc1ccc(-c2nnc(N3CCN(c4ccccn4)CC3)o2)cc1', 'CCN(C)C(=O)Oc1ccc(-c2cc(=O)c3c(O)cc(OC(=O)N(C)CC)cc3o2)cc1', 'CCN(C)C(=O)Oc1ccc(-c2cc(=O)c3c(O)cc(OC(=O)N(C)CC)cc3o2)cc1', 'COc1ccc2[nH]cc(CCNC(=O)CN(CCCCCCCCNc3c4c(nc5ccccc35)CCCC4)C(=O)CCCC[C@@H]3CCSS3)c2c1', 'O=C1c2ccccc2C(=O)C12N1CSC[C@@H]1[C@@H](c1ccc(Br)cc1)[C@@]21CSc2ccccc2C1=O', 'COc1cc2c(cc1O)CCC(C)(CCC(=O)NCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'Cc1cc(S(=O)(=O)NN)c(C(C)C)cc1O', 'COc1ccc(COc2c(OC)ccc3cc4[n+](cc23)CCc2cc3c(cc2-4)OCO3)cc1.[Br-]', 'CCN(Cc1ccccc1)C(=O)/C=C/c1ccc(OC(=O)N(c2ccccc2)c2ccccc2)c(OC)c1', 'COc1cc(/C=C/C(=O)N2CCN(Cc3ccccc3)CC2)ccc1OC(=O)N(c1ccccc1)c1ccccc1', 'c1ccc(-c2nnc(Nc3ncccn3)o2)cc1', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(N)cc3)CC2)nc1-c1ccc(OC)cc1', 'O=C(NO)c1ccc(Nc2c3ccccc3nc3ccccc23)cc1', 'COc1cccc(COc2c(OC)ccc3cc4[n+](cc23)CCc2cc3c(cc2-4)OCO3)c1.[Br-]', 'O=C(N/N=C/c1ccc(F)cc1F)N1CCN(c2ccccn2)CC1', 'CN1C[C@@H](c2cccs2)[C@@]2(CSc3ccccc3C2=O)C12C(=O)c1ccccc1C2=O', 'Cc1ccc(CNC2CCN(Cc3ccccc3)C2)cc1', 'O=C(CCCCCN1CCC2c3ccccc3NC2C1)Nc1ccc(/C=C2\\CCn3c2nc2ccccc2c3=O)cc1', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(N)c(OC)c3)CC2)nc1-c1ccccc1', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(NC(C)=O)cc3)CC2)nc1-c1ccc(OC)c(OC)c1', 'Brc1ccc(-c2nnc(Nc3ncccn3)o2)cc1', 'Oc1ccc(CNC2CCN(Cc3ccccc3)C2)cc1', 'NC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCC1CCN(Cc2ccc(F)cc2)CC1', 'FC(F)(F)c1ccc(-c2nnc(N3CCN(c4ccccn4)CC3)o2)cc1', 'C[C@@H]1CC[C@H]2[C@H](C(=O)c3c4c(cn(O)c3=O)[C@@]3(O)CC[C@H](O)[C@H](O4)[C@@H]3O)[C@H](C)C=C[C@@H]2C1', 'CCn1cc(CNCc2c[nH]nc2-c2ccc(-c3ccccc3)cc2)cn1', 'CCn1cc(CNCc2c[nH]nc2-c2ccc(-c3ccccc3)cc2)cn1', 'CC(C)c1ccc(NC(=O)OC2CCN(C3Cc4ccccc4C3)CC2)cc1', 'COc1cccc(C(=O)N2CCN(CC(=O)N3CCN(CCCc4c[nH]c5ccc(F)cc45)CC3)CC2)c1', 'CS[C@@H](Cn1nnc2cc(C#Cc3ccc(-n4ccnc4CO)cc3)ccc21)N(O)C=O', 'O=c1ccoc2ccc(OCCCCN3CCCCC3)cc12', 'O=c1ccoc2ccc(OCCCCN3CCCCC3)cc12', 'COc1cccc2cc[n+](-c3ccccc3F)cc12.[Br-]', 'COc1cccc2cc[n+](-c3ccc(Br)cc3F)cc12.[Br-]', 'NC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCCC1CCN(Cc2cccc(F)c2)CC1', 'O=C(CN1CCC2c3ccccc3NC2C1)Oc1ccc(/C=C2\\CCn3c2nc2ccccc2c3=O)cc1', 'CCN(C)C(=O)Oc1ccc(/C=C/C(=O)N2CCC(Cc3ccccc3)CC2)cc1', 'CCCC[N+]1(C)CCN(c2ccc(/C=C3\\Cc4cc(OC)c(OC)cc4C3=O)cc2)CC1.[Br-]', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(NC(C)=O)c(OC)c3)CC2)nc1-c1ccccc1', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2cccc(Br)c2)ccc1O', 'CCN(C)C(=O)Oc1ccc(/C=C/C(=O)N2CCc3ccccc3C2)cc1', 'COc1ccc(-c2nnc(Nc3ncccn3)o2)cc1OC', 'O=C(CN1CCN(C(=O)c2ccccc2Br)CC1)N1CCN(CCCc2c[nH]c3ccc(F)cc23)CC1', 'CNC(=O)c1sc(-c2ccnc(NC(=O)CNCCNc3c4c(nc5cc(Cl)ccc35)CCCC4)c2)nc1OCC1CC1', 'O=c1cc(-c2ccc(OCCCCN3CCN(Cc4ccccc4)CC3)cc2)oc2cc(OCCCCN3CCN(Cc4ccccc4)CC3)cc(O)c12', 'O=C(/C=C/c1ccc(OCCCCN2CCCCC2)cc1)c1c(O)cc(OCCCCN2CCCCC2)cc1O', 'CCN(CC)c1cc(/C=C2/OC(=O)c3cc(OC)c(OC)cc32)ccc1O', 'Oc1ccccc1CNC1CCN(Cc2ccccc2)C1', 'COc1cc2c(cc1OC)/C(=C\\c1ccc(N3CCCCC3)c(O)c1)OC2=O', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2ccc(Cl)cc2)ccc1O', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(N3CC[N+](C)(Cc4cccc(Br)c4)CC3)cc1)C2.[Br-]', 'O=c1ccoc2ccc(OCCCN3CCCCC3)cc12', 'COc1cccc2cc[n+](-c3ccccc3I)cc12.[Br-]', 'CCc1ccc(NC(=O)COc2c(Cl)cc(Cl)cc2S(=O)(=O)N2CCN(C(=O)c3ccco3)CC2)cc1', 'O=c1ccoc2ccc(OCCCCCN3CCCC3)cc12', 'O=c1cc(-c2ccc(OCCCCN3CCC(Cc4ccccc4)CC3)cc2)oc2cc(OCCCCN3CCC(Cc4ccccc4)CC3)cc(O)c12', 'COc1cc(/C=C/C(=O)N2CCN(c3ccc(F)cc3)CC2)ccc1O', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(N)c(F)c3)CC2)nc1-c1ccc(OC)cc1', 'Brc1ccc([N+]2=Cc3ccccc3CC2)cc1.[Br-]', 'CNC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCOCc1cn(CCNc2c3c(nc4ccccc24)CCCC3)nn1', 'CCN(CC)C(=O)Oc1ccc(/C=C/C(=O)N2CCc3ccccc3C2)cc1OC', 'Cc1cc(/C=N/O)c(O)c(CNc2ncccc2OCc2ccccc2)c1', 'Oc1ccc(-c2nnc(Nc3ncccn3)o2)cc1', 'Brc1ccc(-c2nnc(N3CCN(c4ccccn4)CC3)o2)cc1', 'COc1cc2c(cc1OC)/C(=C/c1ccc(N3CCCCC3)c(O)c1)OC2=O', 'CN1C[C@@H](c2ccc(Cl)cc2Cl)[C@@]2(CSc3ccccc3C2=O)C12C(=O)c1ccccc1C2=O', 'CS[C@@H](Cn1nnc2cc(C#Cc3ccc(N4C[C@H](O)[C@@H](O)C4)cc3)ccc21)N(O)C=O', 'Oc1ccc(/C=N\\C2CCN(Cc3ccccc3)C2)c(O)c1', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(N3CC[N+](C)(Cc4ccc(Br)cc4)CC3)cc1)C2.[Br-]', 'O=c1ccoc2ccc(OCCCCN3CCCC3)cc12', 'O=C(Nc1ccccc1)OCC1CCN(Cc2ccccc2)CC1', 'O=C(NCCCNc1c2c(nc3ccccc13)CCCC2)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCC1CC1', 'O=C(CCCCN1CCC2c3ccccc3NC2C1)Nc1ccc(/C=C2\\CCn3c2nc2ccccc2c3=O)cc1', 'O=C(CN1CCN(C(=O)c2cccc(Cl)c2)CC1)N1CCN(CCCc2c[nH]c3ccc(F)cc23)CC1', 'COc1cc(/C=C/C(=O)N2CCN(c3ccc(C)cc3)CC2)ccc1O', 'O=c1ccoc2ccc(OCCCCCN3CCOCC3)cc12', 'COc1cccc2cc[n+](-c3ccc(Cl)cc3)cc12.[Br-]', 'Cc1ccc(C(=O)N2CCN(CC(=O)N3CCN(CCCc4c[nH]c5ccc(F)cc45)CC3)CC2)cc1', 'O=C(N/N=C/c1ccc(OC(F)(F)F)cc1)N1CCN(c2ccccn2)CC1', 'Oc1ccccc1-c1nnc(N2CCN(c3ccccn3)CC2)o1', 'Cc1ccc(C(=O)N2CCN(CC(=O)N3CCN(CCCc4c[nH]c5ccccc45)CC3)CC2)cc1', 'COc1cc(CNC(=O)c2ccc(CNc3c4c(nc5cc(Cl)ccc35)CC3C=C(C)CC4C3)cc2)ccc1O', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(NC(C)=O)c(Cl)c3)CC2)nc1-c1ccc(OC)cc1', 'CCN(CCCCOc1ccc(/C=C/C(=O)c2c(O)cc(OCCCCN(CC)Cc3ccccc3OC)cc2O)cc1)Cc1ccccc1OC', 'COc1ccc(C(=O)N2CCN(CC(=O)N3CCN(CCCc4c[nH]c5ccccc45)CC3)CC2)cc1', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2ccc(Br)cc2)ccc1O', 'CCN(CC)CCCCOc1ccc(/C=C/C(=O)c2c(O)cc(OCCCCN(CC)CC)cc2O)cc1', 'COc1cc(CNC(=O)CCCCCCCCNc2c3c(nc4cc(Cl)ccc24)CC2C=C(C)CC3C2)ccc1O', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2ccccc2Cl)ccc1O', 'COc1ccc2[nH]cc(CCCN3CCN(C(=O)CN4CCN(C(=O)c5cccc(F)c5)CC4)CC3)c2c1', 'CCN(CCCCOc1ccc(-c2cc(=O)c3c(O)cc(OCCCCN(CC)Cc4ccccc4)cc3o2)cc1)Cc1ccccc1', 'CCN(CCCCOc1ccc(-c2cc(=O)c3c(O)cc(OCCCCN(CC)Cc4ccccc4)cc3o2)cc1)Cc1ccccc1', 'O=c1c2cc(Cl)ccc2oc2ccc(OCCCCCN3CCCCCC3)cc12', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(N3CC[N+](C)(Cc4ccccn4)CC3)cc1)C2.[Cl-]', 'O=C(CN1CCN(C(=O)c2ccc(F)cc2)CC1)N1CCN(CCCc2c[nH]c3ccccc23)CC1', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2ccccc2OC)ccc1O', 'COc1cc(-c2nnc(Nc3ncccn3)o2)cc(OC)c1OC', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2ccc(C#N)cc2)ccc1O', 'COc1cccc2cc[n+](-c3cccc(C)c3)cc12.[Br-]', 'COc1cccc(-[n+]2ccc3cccc(OC)c3c2)c1.[Br-]', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2ccc(C)cc2)ccc1O', 'COc1ccc2nc3c(c(NCCCCCCCN(CC(=O)NCCc4c[nH]c5ccc(OC)cc45)C(=O)/C=C/c4ccc(O)c(OC)c4)c2c1)CCCC3', 'COc1cccc2cc[n+](-c3cccc(C#N)c3)cc12.[Br-]', 'CNC(=O)c1sc(-c2ccnc(NC(=O)CNCCNc3c4c(nc5ccc(OC)cc35)CCCC4)c2)nc1OCC1CC1', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(NC(C)=O)c(OC)c3)CC2)nc1-c1ccc(OC)c(OC)c1', 'COc1cc2c(cc1O)CCC(C)(CCNCCCCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2cccc(C(F)(F)F)c2)ccc1O', 'COc1cccc(N2CCN(C(=O)/C=C/c3ccc(O)c(OC)c3)CC2)c1', 'COc1cccc(N2CCN(C(=O)/C=C/c3ccc(O)c(OC)c3)CC2)c1', 'COc1cccc2cc[n+](-c3ccccc3C#N)cc12.[Br-]', 'CC(C)c1ccc(NC(=O)OCC2CCCN(C3Cc4ccccc4C3)C2)cc1', 'COc1cc(/C=C/C(=O)N2CCN(c3cccc([N+](=O)[O-])c3)CC2)ccc1O', 'Cc1cc(S(=O)(=O)N2CCOCC2)c(C(C)C)cc1O', 'Clc1ccc(/C=N\\C2CCN(Cc3ccccc3)C2)cc1', 'O=C(CN1CCN(C(=O)c2ccc(F)cc2)CC1)N1CCN(CCCc2c[nH]c3ccc(F)cc23)CC1', 'COc1cccc2cc[n+](-c3ccc(Cl)cc3Cl)cc12.[Br-]', 'O=C(CN1CCN(C(=O)c2cccc(Br)c2)CC1)N1CCN(CCCc2c[nH]c3ccc(F)cc23)CC1', 'COc1cccc2cc[n+](-c3ccccc3Cl)cc12.[Br-]', 'O=C(CN1CCN(C(=O)c2cccc(Cl)c2)CC1)N1CCN(CCCc2c[nH]c3ccccc23)CC1', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(N3CCN(C)CC3)cc1)C2', 'COc1cc(/C=C/C(=O)N2CCN(c3ccccc3)CC2)ccc1O', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2cccc(Cl)c2)ccc1O', 'CC(C)c1ccc(NC(=O)OCC2CCN(C3Cc4ccccc4C3)CC2)cc1', 'CCN(C)C(=O)Oc1ccc(/C=C/C(=O)N2CCC(Cc3ccccc3)CC2)cc1OC', 'CNC(=O)c1sc(-c2ccnc(NC(=O)CNCCNc3c4c(nc5ccccc35)CCCC4)c2)nc1OC(C)C', 'COc1cc2c(cc1O)CCC(C)(CCNCCOCCOCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'CCCCN[C@@H](Cc1c[nH]c2ccccc12)C(=O)NCCC1CCN(Cc2ccccc2)CC1', 'CCN(C)C(=O)Oc1ccc(-c2cc(=O)c3c(OC(=O)N(C)CC)cc(OC(=O)N(C)CC)cc3o2)cc1', 'CCN(C)C(=O)Oc1ccc(-c2cc(=O)c3c(OC(=O)N(C)CC)cc(OC(=O)N(C)CC)cc3o2)cc1', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2ccccc2C(F)(F)F)ccc1O', 'COc1cc2c(cc1O)CCC(C)(CC(=O)NCCCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'O=c1c(-c2ccc(OCCCCN3CCCCC3)cc2)coc2cc(OCCCCN3CCCCC3)cc(O)c12', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2ccc(F)cc2)ccc1O', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(N3CC[N+](C)(Cc4ccccc4Cl)CC3)cc1)C2.[Cl-]', 'COc1cc(/C=C/C(=O)N2CCN(c3ccc(Cl)cc3)CC2)ccc1O', 'O=C(N/N=C/c1ccc(F)cc1)N1CCN(c2ccccn2)CC1', 'O=C1c2ccccc2C(=O)C12N1CSC[C@@H]1[C@@H](c1ccc(F)cc1)[C@@]21CSc2ccccc2C1=O', 'COc1cc(CNC(=O)CCCCCNc2c3c(nc4cc(Cl)ccc24)CC2C=C(C)CC3C2)ccc1O', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2ccccc2)ccc1O', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2ccccc2)ccc1O', 'COc1cccc2cc[n+](-c3c(F)cccc3F)cc12.[Br-]', 'O=C(Nc1ccccc1)OCC1CCCN(Cc2ccccc2)C1', 'COc1cccc2cc[n+](-c3cccc(Cl)c3)cc12.[Br-]', 'COc1cccc2cc[n+](-c3ccc(Br)cc3)cc12.[Br-]', 'COc1ccc2cc3[n+](cc2c1OCc1cc(C)cc(C)c1)CCc1cc2c(cc1-3)OCO2.[Br-]', 'CON(C)C(=O)Oc1ccc(-c2cc(=O)c3c(O)cc(OC(=O)N(C)OC)cc3o2)cc1', 'CNC(=O)c1sc(-c2ccnc(NC(=O)CNCCNc3c4c(nc5ccccc35)CCCC4)c2)nc1OCC1CC1', 'COc1cccc2cc[n+](-c3ccccc3)cc12.[Br-]', 'COc1cc2c(cc1OCc1ccccc1)CCC(C)(CC(=O)NCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'COc1cccc2cc[n+](-c3ccc(Br)cc3Br)cc12.[Br-]', 'CC(C)c1ccc(NC(=O)OC2CCCN(C3Cc4ccccc4C3)C2)cc1', 'COc1ccc(NC(=O)CNC(=O)/C=C/c2ccc(O)c(OC)c2)cc1', 'O=C(CN1CCN(C(=O)c2ccc(Cl)cc2)CC1)N1CCN(CCCc2c[nH]c3ccccc23)CC1', 'COc1cc2c(cc1OCc1ccccc1)CCC(C)(CC(=O)NCCCCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2ccc3ncccc3c2)ccc1O', 'O=C1c2ccccc2C(=O)C12N1CSC[C@@H]1[C@@H](c1cccs1)[C@@]21CSc2ccccc2C1=O', 'CCN(CC)C(=O)/C=C/c1ccc(OC(=O)N(C)C)c(OC)c1', 'C#CCOc1ccc(-c2cc(-c3ccc(OCCN4CCOCC4)cc3)nc(N)n2)cc1', 'O=C(CN1CCN(C(=O)c2ccccc2F)CC1)N1CCN(CCCc2c[nH]c3ccccc23)CC1', 'COc1ccc2cc(C(=O)NCCC3CCN(Cc4ccccc4)CC3)ccc2c1', 'COc1ccc2cc(C(=O)NCCC3CCN(Cc4ccccc4)CC3)ccc2c1', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(N3CC[N+](C)(CCO)CC3)cc1)C2.[Cl-]', 'COc1ccc2cc3[n+](cc2c1OCc1ccc(SC)cc1)CCc1cc2c(cc1-3)OCO2.[Br-]', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(NC(C)=O)c(Cl)c3)CC2)nc1-c1ccc(OC)c(OC)c1', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(NC(C)=O)c(Cl)c3)CC2)nc1-c1ccc(OC)c(OC)c1', 'COc1cccc2cc[n+](-c3ccc(I)cc3F)cc12.[Br-]', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(N3CC[N+](C)(Cc4cccc(Cl)c4)CC3)cc1)C2.[Br-]', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(N3CC[N+](C)(Cc4cccc(Cl)c4)CC3)cc1)C2.[Br-]', 'CC(C)c1ccc([C@@H]2CN(C)C3(C(=O)c4ccccc4C3=O)[C@]23CSc2ccccc2C3=O)cc1', 'CNC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCC1CCN(Cc2ccc(Cl)cc2)CC1', 'CN1C[C@@H](c2ccc(F)cc2)[C@@]2(CSc3ccccc3C2=O)C12C(=O)c1ccccc1C2=O', 'O=c1c(-c2ccc(OCCCCN3CCc4ccccc4C3)cc2)coc2cc(OCCCCN3CCc4ccccc4C3)cc(O)c12', 'O=C(CCCCN1CCC2c3ccccc3NC2C1)Oc1ccc(/C=C2\\CCn3c2nc2ccccc2c3=O)cc1', 'COc1ccc(N2CCN(C(=O)/C=C/c3ccc(O)c(OC)c3)CC2)cc1', 'Clc1ccc(-c2nnc(Nc3ncccn3)o2)cc1', 'COc1cc(/C=C/C(=O)N2CCC(Cc3ccccc3)CC2)ccc1OC(=O)N(C)C', 'CC(C)c1ccc(NC(=O)OCC2CCCN(Cc3ccccc3)C2)cc1', 'O=c1cc(-c2ccc(OCCCCN3CCc4ccccc4C3)cc2)oc2cc(OCCCCN3CCc4ccccc4C3)cc(O)c12', 'CC1c2ccccc2CCN1CCCCOc1ccc(-c2coc3cc(OCCCCN4CCc5ccccc5C4C)cc(O)c3c2=O)cc1', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(N)c(F)c3)CC2)nc1-c1ccc(OC)c(OC)c1', 'CN1CCCC[C@H]1c1ccc(-c2cccnc2)nc1', 'COc1cc2c(cc1O)CCC(C)(CC(=O)NCCCCCCCCNc1c3c(nc4ccccc14)CCCC3)O2', 'CCN(C)C(=O)OCC1CCCN(Cc2ccccc2)C1', 'CNC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCC1CCN(Cc2cccc([N+](=O)[O-])c2)CC1', 'FC(F)(F)c1ccc(/C=N\\C2CCN(Cc3ccccc3)C2)cc1', 'O=C(N/N=C/c1ccc(O)cc1)N1CCN(c2ccccn2)CC1', 'CCN(C)C(=O)Oc1ccc(/C=C/C(=O)N2CCc3ccccc3C2)cc1OC', 'O=C(/C=C/c1ccc(OCCCCN2CCC(Cc3ccccc3)CC2)cc1)c1c(O)cc(OCCCCN2CCC(Cc3ccccc3)CC2)cc1O', 'CS(=O)(=O)[O-].CS(=O)(=O)[O-].NC(=O)c1cc[n+](COC[n+]2ccccc2/C=N/O)cc1', 'CS(=O)(=O)[O-].CS(=O)(=O)[O-].NC(=O)c1cc[n+](COC[n+]2ccccc2/C=N/O)cc1', 'FC(F)(F)Oc1ccc(CNC2CCN(Cc3ccccc3)C2)cc1', 'FC(F)(F)Oc1ccc(CNC2CCN(Cc3ccccc3)C2)cc1', 'FC(F)(F)Oc1ccc(-c2nnc(N3CCN(c4ccccn4)CC3)o2)cc1', 'COc1cccc2cc[n+](-c3ccc(I)cc3)cc12.[Br-]', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2cccc(C#N)c2)ccc1O', 'COc1cccc2cc[n+](-c3cccc(F)c3)cc12.[Br-]', 'Brc1ccc(/C=N\\C2CCN(Cc3ccccc3)C2)cc1', 'O=C1CC(c2ccc(OCCCCN3CCc4ccccc4C3)cc2)Oc2cc(OCCCCN3CCc4ccccc4C3)cc(O)c21', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(N3CC[N+](C)(Cc4ccccc4Br)CC3)cc1)C2.[Br-]', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2ccc(C(F)(F)F)cc2)ccc1O', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(OCCN[N+]3(C)CCN(c4ccc(/C=C5\\Cc6cc(OC)c(OC)cc6C5=O)cc4)CC3)cc1)C2.[Br-]', 'CCCCCC[N+]1(C)CCN(c2ccc(/C=C3\\Cc4cc(OC)c(OC)cc4C3=O)cc2)CC1.[Br-]', 'COc1cc2c(cc1O)CCC(C)(CC(=O)NCCCCCCCNc1c3c(nc4ccccc14)CCCC3)O2', 'COc1cc2c(cc1O)CCC(C)(CCC(=O)NCCCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'CCC[N+]1(C)CCN(c2ccc(/C=C3\\Cc4cc(OC)c(OC)cc4C3=O)cc2)CC1.[Br-]', 'COc1cc(CNC(=O)CCCC/C=C/CCCNc2c3c(nc4cc(Cl)ccc24)CC2C=C(C)CC3C2)ccc1O', 'CN1C[C@@H](c2cccc3ccccc23)[C@@]2(CSc3ccccc3C2=O)C12C(=O)c1ccccc1C2=O', 'NC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCCC1CCN(Cc2ccccc2)CC1', 'O=C1CC(c2ccc(OCCCCN3CCN(Cc4ccccc4)CC3)cc2)Oc2cc(OCCCCN3CCN(Cc4ccccc4)CC3)cc(O)c21', 'O=C(CN1CCN(C(=O)c2ccc(Br)cc2)CC1)N1CCN(CCCc2c[nH]c3ccc(F)cc23)CC1', 'CCN(CCCCOc1ccc(C2CC(=O)c3c(O)cc(OCCCCN(CC)Cc4ccccc4OC)cc3O2)cc1)Cc1ccccc1OC', 'CON(C)C(=O)Oc1ccc(-c2cc(=O)c3c(OC(=O)N(C)OC)cc(OC(=O)N(C)OC)cc3o2)cc1', 'CC(C)c1ccc(NC(=O)OCC2CCN(Cc3ccccc3)CC2)cc1', 'COc1ccc2cc3[n+](cc2c1OCc1ccc(C#N)cc1)CCc1cc2c(cc1-3)OCO2.[Br-]', 'NC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCC1CCN(Cc2cccc(F)c2)CC1', 'Cn1c(=O)sc2cc(C(=O)CCN3CCN(Cc4ccccc4)CC3)ccc21', 'O=C(NO)c1ccc(Nc2c3ccccc3nc3ccc(Br)cc23)cc1', 'COc1cccc2cc[n+](-c3ccccc3Br)cc12.[Br-]', 'C(=N\\C1CCN(Cc2ccccc2)C1)\\c1ccccc1', 'O=C(CN1CCN(C(=O)c2ccc(Cl)cc2)CC1)N1CCN(CCCc2c[nH]c3ccc(F)cc23)CC1', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2ccccc2F)ccc1O', 'NC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCC1CCN(Cc2ccccc2)CC1', 'COc1cc2c(cc1O)CCC(C)(CCNCCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'Brc1ccc(-[n+]2ccc3ccccc3c2)cc1.[Br-]', 'COc1cccc2cc[n+](-c3ccc(F)cc3)cc12.[Br-]', 'O=c1c(-c2ccc(OCCCCN3CCC(Cc4ccccc4)CC3)cc2)coc2cc(OCCCCN3CCC(Cc4ccccc4)CC3)cc(O)c12', 'CNC(=O)c1sc(-c2ccnc(NC(=O)C3CC3)c2)nc1OCCc1cn(CCNc2c3c(nc4ccccc24)CCCC3)nn1', 'O=c1c(-c2ccc(OCCCCN3CCN(Cc4ccccc4)CC3)cc2)coc2cc(OCCCCN3CCN(Cc4ccccc4)CC3)cc(O)c12', 'O=[N+]([O-])c1ccc(CNC2CCN(Cc3ccccc3)C2)cc1', 'Cc1ccc([C@@H]2CN(C)C3(C(=O)c4ccccc4C3=O)[C@]23CSc2ccccc2C3=O)cc1', 'Fc1ccc(/C=N\\C2CCN(Cc3ccccc3)C2)cc1', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2cccc(C)c2)ccc1O', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2ccccc2C)ccc1O', 'COc1cc2c(cc1O)CCC(C)(CCC(=O)NCCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(N)c(OC)c3)CC2)nc1-c1ccc(OC)cc1', 'O=c1c2ccccc2oc2ccc(OCCCCCN3CCCCCC3)cc12', 'O=C1c2ccccc2C(=O)C12N1CSC[C@@H]1[C@@H](c1cccc(Br)c1)[C@@]21CSc2ccccc2C1=O', 'Cc1cc(S(=O)(=O)NCc2ccccn2)c(C(C)C)cc1O', 'COc1ccc(-[n+]2ccc3cccc(OC)c3c2)cc1.[Br-]', 'CCCCC[N+]1(C)CCN(c2ccc(/C=C3\\Cc4cc(OC)c(OC)cc4C3=O)cc2)CC1.[Br-]', 'COc1cc2c(cc1O)CCC(C)(CC(=O)NCCCCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'O=C1c2ccccc2C(=O)C12N1CSC[C@@H]1[C@@H](c1ccc(Cl)cc1Cl)[C@@]21CSc2ccccc2C1=O', 'Cc1ccc(-c2nnc(Nc3ncccn3)o2)cc1', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(NC(C)=O)c(F)c3)CC2)nc1-c1ccc(OC)cc1', 'C=CC[N+]1(C)CCN(c2ccc(/C=C3\\Cc4cc(OC)c(OC)cc4C3=O)cc2)CC1.[Cl-]', 'O=C(CN1CCN(C(=O)c2cccc(Br)c2)CC1)N1CCN(CCCc2c[nH]c3ccccc23)CC1', 'CNC(=O)c1sc(-c2ccnc(NC(=O)CNCCCCCCCCNc3c4c(nc5ccccc35)CCCC4)c2)nc1OCC1CC1', 'O=C(/C=C/c1ccc(OCCCCN2CCCC2)cc1)c1c(O)cc(OCCCCN2CCCC2)cc1O', 'O=C1CC(c2ccc(OCCCCN3CCC(Cc4ccccc4)CC3)cc2)Oc2cc(OCCCCN3CCC(Cc4ccccc4)CC3)cc(O)c21', 'COc1cc2c(cc1O)CCC(C)(CC(=O)NCCCCCCCNc1c3c(nc4cc(Cl)ccc14)CCCC3)O2', 'COc1cccc2cc[n+](-c3ccccc3C)cc12.[Br-]', 'Cc1ccc(/C=N/NC(=O)N2CCN(c3ccccn3)CC2)cc1', 'CCN(C)C(=O)OCC1CCN(Cc2ccccc2)CC1', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2cccc(F)c2)ccc1O', 'CN1C[C@@H](c2cccc(F)c2)[C@@]2(CSc3ccccc3C2=O)C12C(=O)c1ccccc1C2=O', 'COc1cccc2cc[n+](-c3ccc(C(F)(F)F)cc3)cc12.[Br-]', 'O=C(Nc1ccccc1)OC1CCN(C2Cc3ccccc3C2)CC1', 'COc1cccc(NC(=O)CNC(=O)/C=C/c2ccc(O)c(OC)c2)c1', 'O=C(NO)c1ccc(Nc2c3c(nc4ccccc24)CCCC3)cc1', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2ccc3[nH]ccc3c2)ccc1O', 'COc1cc(/C=C/C(=O)NCC(=O)Nc2ccc3[nH]ccc3c2)ccc1O', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(N)c(Cl)c3)CC2)nc1-c1ccc(OC)c(OC)c1', 'CN1C[C@@H](c2ccccc2)[C@@]2(CSc3ccccc3C2=O)C12C(=O)c1ccccc1C2=O', 'C[C@H]1CCCN1Cc1nc2cc(NC(=O)c3ccc4cnccc4c3)ccc2[nH]1', 'COc1ccc(C(=O)N2CCN(CC(=O)N3CCN(CCCc4c[nH]c5ccc(F)cc45)CC3)CC2)cc1', 'COc1cc(/C=N/NC(=O)N2CCN(c3ccccn3)CC2)cc(OC)c1OC', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(N3CC[N+](C)(Cc4ccccc4)CC3)cc1)C2.[Cl-]', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1ccc(N3CC[N+](C)(Cc4ccccc4)CC3)cc1)C2.[Cl-]', 'Cc1cc(S(=O)(=O)Nc2ccccc2O)c(C(C)C)cc1O', 'CCN(C)C(=O)Oc1ccc(/C=C/C(=O)N(CC)Cc2ccccc2)cc1OC', 'COc1cccc2cc[n+](-c3ccc(C)cc3)cc12.[Br-]', 'CNC(=O)c1sc(-c2ccnc(NC(=O)CNCCCCCCNc3c4c(nc5ccccc35)CCCC4)c2)nc1OCC1CC1', 'CCOC(=O)c1c(C)nc(N2CCN(S(=O)(=O)c3ccc(N)c(Cl)c3)CC2)nc1-c1ccccc1', 'O=C(CN1CCN(C(=O)c2ccc(Br)cc2)CC1)N1CCN(CCCc2c[nH]c3ccccc23)CC1', 'Cc1ccc2c(N)c3c(nc2c1)CCCCC3.Cl', 'Cc1ccc(S(=O)(=O)SCc2ccc(F)cc2)cc1', 'Cc1ccc(S(=O)(=O)SCc2ccc(F)cc2)cc1', 'COc1ccc2[nH]cc(CCNC(=O)CCCCn3cc(CN4CC[C@@]56C=C[C@H](O)C[C@@H]5Oc5c(OC)ccc(c56)C4)nn3)c2c1', 'COc1ccc(CN2CCC(c3oc4ccccc4c3C(=O)c3ccc(OC)cc3)CC2)cc1', 'COc1cc2c(cc1OC)C(=O)C(CC1(F)CCN(Cc3ccccc3)CC1)C2', 'O=C(CC1CCN(Cc2ccccc2F)CC1)Nc1n[nH]c2cc(-c3cccnc3)ccc12', 'COc1ccc2[nH]c3ccncc3c2c1', 'CCOC(=O)/C(=N\\Nc1ccccc1C(=O)OCC)C(=O)C(F)(F)F', 'COc1ccc(CN2CCC(c3oc4ccccc4c3C(=O)c3ccccc3)CC2)cc1OC', 'O=C(CCn1c2ccccc2c2cnccc21)N1c2ccccc2Sc2ccccc21', 'Cl.Nc1c2c(nc3ccc(F)cc13)CCCC2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCC(=O)NCc4cc[nH]c(=O)c4)c3C(C1)C2', 'COc1cccc(NC(=O)CCNC(=O)/C=C/c2ccc(O)c(OC)c2)c1', 'COc1cc(/C=C/C(=O)N2CCN(Cc3ccccc3)CC2)ccc1O', 'COc1ccc(OC)c(CN2CCC(c3oc4ccccc4c3C(=O)c3ccc(OC)c(OC)c3)CC2)c1', 'COc1ccc(C(=O)c2c(C3CCN(Cc4ccccc4)CC3)oc3ccccc23)cc1', 'C[n+]1ccc2[nH]c3ccc(F)cc3c2c1.[I-]', 'CCOC(=O)/C(=N\\Nc1ccc(C)cc1)C(=O)C(F)(F)F', 'COc1ccc2[nH]cc(CCNC(=O)CCCCn3cc(CCCCN4CC[C@@]56C=C[C@H](O)C[C@@H]5Oc5c(OC)ccc(c56)C4)nn3)c2c1', 'COC(=O)/C(=N\\Nc1ccccc1)C(=O)C(F)(F)F', 'COc1ccc2[nH]c3c(c2c1)CN(Cc1ccccc1)CC3', 'COc1ccc2[nH]cc(CCNC(=O)Cn3cc(CN4CCc5cc(OC)c(OC)cc5C4CCCN4C(=O)c5ccccc5C4=O)nn3)c2c1', 'O=C(NC1CCN(Cc2ccc(F)cc2)CC1)c1cncc(NC(=O)c2ccnc3[nH]ccc23)c1', 'c1ccc(CN2CCC(CCNc3cc4c(nn3)-c3ccccc3CCCC4)CC2)cc1', '[Br-].c1ccc(C[n+]2ccc3[nH]c4ccccc4c3c2)cc1', 'COc1ccc2c(c1)c1cnccc1n2CCC(=O)N1c2ccccc2Sc2ccccc21', 'COc1ccc2nc3c(c(N)c2c1)C1CC(C)=CC(C3)C1.Cl', 'O=C(CCn1c2ccccc2c2c[n+](Cc3ccccc3)ccc21)N1c2ccccc2Sc2ccccc21.[Br-]', 'COc1cccc(CN2CCN(C(=O)/C=C/c3ccc(O)c(OC)c3)CC2)c1', 'Fc1ccccc1-c1ccc(N[C@H]2C[C@@H]3CN(CC4CCOCC4)C[C@@H]3C2)nn1', 'O=C(NCC1CCN(Cc2ccccc2)CC1)c1cncc(NC(=O)c2ccnc(NC(=O)C3CC3)c2)c1', 'COc1ccc2[nH]cc(CCNC(=O)Cn3cc(CCCCN4CC[C@@]56C=C[C@H](O)C[C@@H]5Oc5c(OC)ccc(c56)C4)nn3)c2c1', 'Cc1ccc(S(=O)(=O)SCc2ccc(C(C)(C)C)cc2)cc1', 'Cc1ccc(S(=O)(=O)SCc2ccc(C(C)(C)C)cc2)cc1', 'COc1ccc(CN2CCC(c3oc4ccccc4c3C(=O)c3ccccc3)CC2)cc1', 'CCOC(=O)/C(=N\\Nc1cccc(C)c1)C(=O)C(F)(F)F', 'O=C(NCCC1CCN(Cc2cccc(Cl)c2)CC1)c1cncc(NC(=O)c2cccc(NC(=O)C3CC3)c2)c1', 'COc1cc2c(cc1OC)C(=O)C(CC1(F)CCN(Cc3cccc(F)c3)CC1)C2', 'CCOC(=O)/C(=N\\Nc1ccccc1)C(=O)C(F)(F)F', 'O=C(Nc1cncc(C(=O)NC2CCN(Cc3cccc(Cl)c3)CC2)c1)c1ccnc(NC(=O)C2CC2)c1', 'CCOC(=O)/C(=N\\Nc1ccccc1[N+](=O)[O-])C(=O)C(F)(F)F', 'COc1cc(/C=C/C(=O)NCC(=O)NCCc2c[nH]c3ccccc23)ccc1O', 'COc1ccc2[nH]cc(CCNC(=O)Cn3cc(CN4CC[C@@]56C=C[C@H](O)C[C@@H]5Oc5c(OC)ccc(c56)C4)nn3)c2c1', 'Cl.Nc1c2c(nc3c(Cl)cc(Cl)cc13)CCCC2', 'O=C(NCC1CCN(Cc2ccccc2F)CC1)c1n[nH]c2ccc(-c3cccnc3)cc12', 'Cl.Nc1c2c(nc3ccc(F)cc13)CCC2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCC(=O)N4CCNC(=O)C4)c3C(C1)C2', 'Cc1ccc2c(c1)c1cnccc1n2CCC(=O)N1c2ccccc2Sc2ccccc21', 'CCc1ccc(CN2CCC(c3cc4ccccc4o3)CC2)cc1', 'O=C(Nc1cccc(OCC2CCN(Cc3ccccc3)CC2)c1)c1ccnc(NC(=O)C2CC2)c1', 'COc1ccc2nc3c(c(N)c2c1)CCC3.Cl', 'c1ccc(CN2CCC(c3cc4ccccc4o3)CC2)cc1', 'CCOC(=O)/C(=N\\Nc1ccc(OC)cc1)C(=O)C(F)(F)F', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(CCCCc1cn(CCCC(=O)NCCc3c[nH]c4ccccc34)nn1)C2', 'COC(=O)/C(=N\\Nc1ccccc1[N+](=O)[O-])C(=O)C(F)(F)F', 'Cc1cccc2nc3c(c(N)c12)CCCCC3.Cl', 'CCOC(=O)/C(=N\\Nc1ccc(C(=O)OCC)cc1)C(=O)C(F)(F)F', 'COc1cc(/C=C/C(=O)NCCC(=O)Nc2ccc(Cl)cc2)ccc1O', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCNC(=O)[C@@H](N)Cc4c[nH]c5ccccc45)c3C(C1)C2.Cl.Cl', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCNC(=O)[C@@H](N)Cc4c[nH]c5ccccc45)c3C(C1)C2.Cl.Cl', 'Cc1ccc2[nH]c3c(c2c1)CN(Cc1ccccc1)CC3', 'COc1ccc(CN2CCN(C(=O)/C=C/c3ccc(O)c(OC)c3)CC2)cc1', 'COc1cc2c(cc1OC)C(=O)C(CC1(F)CCN(Cc3cc(F)c(F)cc3F)CC1)C2', 'Cl.Nc1c2c(nc3ccc(Cl)cc13)CCC2', 'COc1ccc(C(=O)c2c(C3CCN(Cc4cc(OC)ccc4OC)CC3)oc3ccccc23)cc1', 'COc1cc2c(cc1OC)C(=O)C(CC1(F)CCN(Cc3ccc(F)cc3)CC1)C2', 'Cl.Nc1c2c(nc3c(Br)cc(Br)cc13)CCCC2', 'O=C(Nc1cncc(C(=O)NC2CCN(Cc3ccccc3)CC2)c1)c1ccnc(NC(=O)C2CC2)c1', 'Cc1ccc2c(N)c3c(nc2c1)CCC3.Cl', 'COc1ccc2[nH]c3cc[n+](C)cc3c2c1.[I-]', 'Cl.Nc1c2c(nc3ccc(Br)cc13)CCC2', 'Cc1ccc(S(=O)(=O)SCc2ccc(C#N)cc2)cc1', 'Cc1ccc(S(=O)(=O)SCc2ccc(C#N)cc2)cc1', 'CCOC(=O)/C(=N\\Nc1ccc(F)cc1)C(=O)C(F)(F)F', 'COc1ccc(OC)c(CN2CCC(c3oc4ccccc4c3C(=O)c3ccccc3)CC2)c1', 'Cc1ccc2c(N)c3c(nc2c1)CCCC3.Cl', 'O=C(CC1CCN(Cc2ccc(F)cc2)CC1)Nc1n[nH]c2cc(-c3cccnc3)ccc12', 'Fc1ccc2[nH]c3c(c2c1)CN(Cc1ccccc1)CC3', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCNC(=O)c4cc(O)cc(O)c4)c3C(C1)C2', 'Cc1cccc2nc3c(c(N)c12)CCC3.Cl', 'COc1cc2c(cc1OC)C(=O)C(CC1(F)CCN(Cc3ccccc3F)CC1)C2', 'COc1cc(/C=C/C(=O)NCCC(=O)Nc2ccccc2)ccc1O', 'CN1CCc2c(c3ccccc3n2CCC(=O)N2c3ccccc3Sc3ccccc32)C1', 'Cc1ccc(S(=O)(=O)SCc2ccc(OC(F)(F)F)cc2)cc1', 'Cc1ccc(S(=O)(=O)SCc2ccc(OC(F)(F)F)cc2)cc1', 'Cl.Nc1c2c(nc3ccc(Cl)cc13)CCCCC2', 'Cc1ccc2nc3c(c(N)c2c1)CCCCC3.Cl', 'Cl.Nc1c2c(nc3cccc(Cl)c13)CCCC2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCNS(=O)(=O)c4cn[nH]c4)c3C(C1)C2', 'O=C(NC1CCN(Cc2ccccc2)CC1)c1cncc(NC(=O)c2ccnc3[nH]ccc23)c1', 'Fc1ccc2[nH]c3ccncc3c2c1', 'Cc1ccc(S(=O)(=O)SCc2ccc(Cl)cc2)cc1', 'Cc1ccc(S(=O)(=O)SCc2ccc(Cl)cc2)cc1', 'c1ccc2oc(C3CCNCC3)cc2c1', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCC(=O)Nc4cn[nH]c4)c3C(C1)C2', 'COc1ccc(CN2CCC(c3cc4ccccc4o3)CC2)cc1OC', 'O=C(Nc1cncc(OCC2CCN(Cc3ccccc3)CC2)c1)c1ccnc(NC(=O)C2CC2)c1', 'O=C(NCC1CCN(Cc2cccc(Cl)c2)CC1)c1cncc(NC(=O)c2ccnc(NC(=O)C3CC3)c2)c1', 'O=C(CCn1c2ccncc2c2cc(F)ccc21)N1c2ccccc2Sc2ccccc21', 'COc1cc(/C=C/C(=O)N2CCN(Cc3ccc(F)cc3)CC2)ccc1O', 'Cl.Nc1c2c(nc3c(Cl)cc(Cl)cc13)CCC2', 'O=C(Nc1cncc(OCC2CCN(Cc3ccccc3F)CC2)c1)c1ccnc(NC(=O)C2CC2)c1', 'O=C(c1ccccc1)c1c(C2CCN(Cc3ccccc3)CC2)oc2ccccc12', 'O=C(NCC1CCN(Cc2ccc(F)c(F)c2)CC1)c1cncc(NC(=O)c2ccnc(NC(=O)C3CC3)c2)c1', 'Cl.Nc1c2c(nc3ccc(F)cc13)CCCCC2', 'Cc1ccc2[nH]c3cc[n+](C)cc3c2c1.[I-]', 'O=C(NCC1CCN(Cc2cccc(F)c2)CC1)c1cncc(NC(=O)c2ccnc(NC(=O)C3CC3)c2)c1', 'COc1cc(/C=C/C(=O)NCC(=O)NCCc2c[nH]c3ccc(Cl)cc23)ccc1O', 'COc1ccc2[nH]cc(CCNC(=O)Cn3cc(CCCCN(C)CCCC4c5cc(OC)c(OC)cc5CCN4C)nn3)c2c1', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCCCNC(=O)c4ccc[nH]c4=O)c3C(C1)C2', 'Cc1ccc(S(=O)(=O)SCc2ccc([N+](=O)[O-])cc2)cc1', 'Cc1ccc(S(=O)(=O)SCc2ccc([N+](=O)[O-])cc2)cc1', 'O=C(Nc1cncc(OCC2CCN(Cc3ccccc3Cl)CC2)c1)c1ccnc(NC(=O)C2CC2)c1', 'Cl.Nc1c2c(nc3ccc(Br)cc13)CCCCC2', 'CCOC(=O)/C(=N\\Nc1ccc(Br)cc1)C(=O)C(F)(F)F', 'Cl.Nc1c2c(nc3c(Br)cc(Br)cc13)CCC2', 'COc1cc(/C=C/C(=O)N2CCN(Cc3cccc(Cl)c3)CC2)ccc1O', 'COC(=O)/C(=N\\Nc1ccc(C)cc1)C(=O)C(F)(F)F', 'COC(=O)/C(=N\\Nc1ccc(C)cc1)C(=O)C(F)(F)F', 'Cc1ccc(S(=O)(=O)SCc2ccc(C(F)(F)F)cc2)cc1', 'Cc1ccc(S(=O)(=O)SCc2ccc(C(F)(F)F)cc2)cc1', 'O=C(NCC1CCN(Cc2ccc(F)cc2)CC1)c1cncc(NC(=O)c2ccnc(NC(=O)C3CC3)c2)c1', 'Cl.Nc1c2c(nc3c(Br)cc(Br)cc13)CCCCC2', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(CCCCc1cn(CCCCC(=O)NCCc3c[nH]c4ccccc34)nn1)C2', 'COc1ccc2[nH]cc(CCNC(=O)CNC(=O)/C=C/c3ccc(O)c(OC)c3)c2c1', 'O=C(NCC1CCN(Cc2cccc(Cl)c2)CC1)c1cncc(NC(=O)c2ccnc3[nH]ccc23)c1', 'C[n+]1ccc2c(c1)c1ccccc1n2CCC(=O)N1c2ccccc2Sc2ccccc21.[I-]', 'O=C(NCC1CCN(Cc2ccccc2F)CC1)c1cncc(NC(=O)c2ccnc(NC(=O)C3CC3)c2)c1', 'O=C(CC1CCN(Cc2ccccc2)CC1)Nc1n[nH]c2cc(-c3cccnc3)ccc12', 'COc1ccc(CN2CCC(c3cc4ccccc4o3)CC2)cc1', 'COc1ccc2c3c1O[C@H]1C[C@@H](O)C=C[C@@]31CCN(CCCCc1cn(CC(=O)NCCc3c[nH]c4ccccc34)nn1)C2', 'Cc1ccc(S(=O)(=O)SCc2ccc(C(C)C)cc2)cc1', 'Cc1ccc(S(=O)(=O)SCc2ccc(C(C)C)cc2)cc1', 'COc1cc(CNC(=O)CCCCCCCCNc2c3c(nc4cc(Cl)ccc24)CC2C=C(C)CC3C2)ccn1', 'Cc1ccccc1CN1CCC(CNC(=O)c2cncc(NC(=O)c3ccnc(NC(=O)C4CC4)c3)c2)CC1', 'COC(=O)/C(=N\\Nc1ccc(OC)cc1)C(=O)C(F)(F)F', 'COc1ccc2nc3c(c(N)c2c1)CCCC3.Cl', 'Cl.Nc1c2c(nc3c(Cl)cc(Cl)cc13)CCCCC2', 'Oc1ccc(CN2CCC(c3cc4ccccc4o3)CC2)cc1', 'O=C(NCC1CCN(Cc2ccccc2)CC1)c1n[nH]c2ccc(-c3cccnc3)cc12', 'COc1cc(/C=C/C(=O)N2CCN(Cc3cccc(F)c3)CC2)ccc1O', 'Cc1cccc2nc3c(c(N)c12)CCCC3.Cl', 'O=C(Nc1cncc(OCC2CCN(Cc3ccc(F)cc3)CC2)c1)c1ccnc(NC(=O)C2CC2)c1', 'Cc1ccc2c(c1)c1c[n+](C)ccc1n2CCC(=O)N1c2ccccc2Sc2ccccc21.[I-]', 'C[n+]1ccc2c(c1)c1cc(F)ccc1n2CCC(=O)N1c2ccccc2Sc2ccccc21.[I-]', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCNC(=O)[C@@H](N)Cc4c[nH]c5ccccc45)c3C(C1)C2.Cl.Cl', 'COc1ccc(CN2CCC(c3oc4ccccc4c3C(=O)c3ccc(OC)c(OC)c3)CC2)cc1OC', 'Cc1ccc2nc3c(c(N)c2c1)CCC3.Cl', 'COc1ccc2nc3c(c(N)c2c1)CCCCC3.Cl', 'Fc1ccc(CN2CCC(c3cc4ccccc4o3)CC2)cc1', 'COc1ccc(OC)c(CN2CCC(c3cc4ccccc4o3)CC2)c1', 'CC(=O)Nc1cc(C(=O)Nc2cncc(OCC3CCN(Cc4ccc(F)cc4)CC3)c2)ccn1', 'COc1ccc(CSS(=O)(=O)c2ccc(C)cc2)cc1', 'COc1ccc(CSS(=O)(=O)c2ccc(C)cc2)cc1', 'Cc1ccc2nc3c(c(N)c2c1)CCCC3.Cl', 'Cl.Nc1c2c(nc3ccc(Cl)cc13)CCCC2', 'O=C(Nc1cncc(OCC2CCN(Cc3ccccc3)CC2)c1)c1ccnc2[nH]ccc12', 'COc1cc2c(cc1OC)C(=O)C(CC1(F)CCN(Cc3ccc(F)cc3F)CC1)C2', 'CC1=CC2Cc3nc4cc(Cl)ccc4c(NCCCCCCCNC(=O)[C@@H](N)Cc4c[nH]c5ccccc45)c3C(C1)C2.Cl.Cl', 'Clc1ccc(CN2CCC(c3cc4ccccc4o3)CC2)cc1', 'Cl.Nc1c2c(nc3ccc(Br)cc13)CCCC2', 'COc1ccc2c(c1)c1c[n+](C)ccc1n2CCC(=O)N1c2ccccc2Sc2ccccc21.[I-]', 'Cc1c(C)c2ccc(OCC3CCCN(CCCC(F)F)C3)cc2oc1=O', 'Oc1cccc(CCCCCCCCNc2c3c(nc4cc(Cl)ccc24)CCCC3)c1', 'Oc1cccc(CCCCCCCCNc2c3c(nc4ccccc24)CCCC3)c1', 'O=C(O[C@H]1CN2Cc3cc4c(cc3[C@]13C=C[C@@H](O)C[C@H]23)OCO4)c1cccc([N+](=O)[O-])c1', 'CO[C@H]1C[C@H]2C(=C[C@@H]1OC(=O)c1ccc(C)c([N+](=O)[O-])c1)[C@H]1CN2Cc2cc3c(cc21)OCO3', 'COc1cc2c(c3c1OCOC3)C[C@H]1c3c(cc4c(c3OC)OCO4)CC[N+]1=C2.O=C([O-])C(F)(F)F', 'COc1cc2c(c3c1OCOC3)C[C@H]1c3c(cc4c(c3OC)OCO4)CC[N+]1=C2.O=C([O-])C(F)(F)F', 'COc1cc(CCCCCCCCNc2c3c(nc4cc(Cl)ccc24)CCCC3)cc(OC)c1', 'COC(=O)c1c(CCCCCCCCNc2c3c(nc4cc(Cl)ccc24)CCCC3)cccc1OC', 'Cc1c(C)c2ccc(OCc3ccc(CN(C)CCCC(F)F)cc3)cc2oc1=O', 'O=C(O)c1c(O)cccc1CCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'CN(CCCO)Cc1ccc(COc2ccc3c(CO)cc(=O)oc3c2)cc1', 'COc1cc2c[n+]3c(cc2c(CO)c1OC)-c1c(cc2c(c1OC)OCO2)CC3', 'COc1cc2c[n+]3c(cc2c(CO)c1OC)-c1c(cc2c(c1OC)OCO2)CC3', 'COC(=O)c1c(CCCCCCCCNc2c3c(nc4ccccc24)CCCC3)cccc1OC', 'Cc1c(C)c2ccc(OCC3CCN(CCCC(F)F)CC3)cc2oc1=O', 'CCN(CCCOc1ccc2c(C)c(C)c(=O)oc2c1)Cc1cccc(F)c1', 'N#Cc1ccc2[nH]cc(CCCCN3CCN(c4ccc(C(=O)CCC5CCN(Cc6ccccc6)CC5)cc4)CC3)c2c1', 'COC(=O)c1c(O)cccc1CCCCCCCCNc1c2c(nc3cc(Cl)ccc13)CCCC2', 'CN(CCCC(F)F)Cc1ccc(COc2ccc3c(CO)cc(=O)oc3c2)cc1', 'CN(CCCC(F)F)Cc1ccc(COc2ccc3c(CO)cc(=O)oc3c2)cc1', 'Cc1c(C)c2ccc(OCC3CCCN(Cc4cccc(F)c4)C3)cc2oc1=O', 'COc1c(-c2ccc(O)cc2)cc2oc3cc(O)c(O)cc3c2c1O', 'CN(Cc1ccccc1)Cc1ccc(COc2ccc3c(C(F)F)cc(=O)oc3c2)cc1', 'Cc1c(C)c2ccc(OCc3cccc(CN(C)CCCC(F)F)c3)cc2oc1=O', 'CO[C@H]1C[C@H]2C(=C[C@@H]1OC(=O)c1cc([N+](=O)[O-])cc([N+](=O)[O-])c1)[C@H]1CN2Cc2cc3c(cc21)OCO3', 'COC(=O)c1c(CCCCCCCCNc2c3c(nc4ccc(OC)cc24)CCCC3)cccc1OC', 'COc1cc2c(c3c1OCOC3)C[C@H]1c3c(cc4c(c3OC)OCO4)CCN1C2', 'COc1cc2c(c3c1OCOC3)C[C@H]1c3c(cc4c(c3OC)OCO4)CCN1C2', 'Cc1c(C)c2ccc(OCC3CCN(Cc4cccc(C(F)(F)F)c4)CC3)cc2oc1=O', 'O=C(O[C@@H]1C=C[C@]23CCN(Cc4cc5c(cc42)OCO5)[C@H]3C1)c1ccccc1[N+](=O)[O-]', 'CO[C@H]1C[C@H]2C(=C[C@@H]1OC(=O)c1ccc(Cl)c([N+](=O)[O-])c1)[C@H]1CN2Cc2cc3c(cc21)OCO3', 'O=C(O)c1c(O)cccc1CCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'COc1cc(CCCCCCCCNc2c3c(nc4ccccc24)CCCC3)cc(OC)c1', 'COc1cccc(CCCCCCCCNc2c3c(nc4cc(Cl)ccc24)CCCC3)c1', 'O=C(O[C@H]1CN2Cc3cc4c(cc3[C@]13C=C[C@@H](O)C[C@H]23)OCO4)c1ccccc1Cl', 'COc1cc2c(c(CO)c1OC)C[C@H]1c3c(cc4c(c3OC)OCO4)CC[N+]1=C2.O=C([O-])C(F)(F)F', 'COc1cc2c(c(CO)c1OC)C[C@H]1c3c(cc4c(c3OC)OCO4)CC[N+]1=C2.O=C([O-])C(F)(F)F', 'CN(Cc1ccc(COc2ccc3c(CO)cc(=O)oc3c2)cc1)Cc1cccc(F)c1', 'COc1ccccc1C(=O)O[C@H]1CN2Cc3cc4c(cc3[C@]13C=C[C@@H](O)C[C@H]23)OCO4', 'Cc1c(C)c2ccc(OCC3CCN(Cc4cccc(F)c4)CC3)cc2oc1=O', 'CCNC(=O)Oc1ccc2c(c1)[C@]1(C)CCOC1O2', 'COc1cccc(CCCCCCCCNc2c3c(nc4ccccc24)CCCC3)c1', 'CO[C@H]1C[C@H]2C(=C[C@@H]1OC(=O)c1cc([N+](=O)[O-])ccc1Cl)[C@H]1CN2Cc2cc3c(cc21)OCO3', 'COC(=O)c1c(O)cccc1CCCCCCCCNc1c2c(nc3ccccc13)CCCC2', 'Cc1c(C)c2ccc(OCCCN(Cc3cccc(F)c3)C(C)C)cc2oc1=O', 'CCCCN(CCOc1cccc2c(OC(=O)N(C)C)cccc12)CCC1CCCCCC1', 'c1ccc2c(NCCN3CCOCC3)c3c(nc2c1)CCCC3', 'COc1cc(/C=C/c2cc(=O)oc3c2ccc2oc(-c4ccc(C)cc4)cc23)cc(OC)c1', 'O=C(COc1ccc(Cl)cc1)NCCN1CCN(Cc2ccccc2)CC1', 'NS(=O)(=O)c1ccc(NCc2cccc(Cl)c2O)cc1', 'CCN(CC)C(=O)c1ccc2c(c1)nc(Cc1ccc(OC(=O)NCCCCCCN3CCc4ccccc4C3)cc1)n2CCC(C)C', 'CCN(CC)C(=O)c1ccc2c(c1)nc(Cc1ccc(OC(=O)NCCCCCCN3CCc4ccccc4C3)cc1)n2CCC(C)C', 'O=C(COc1ccc(-c2ccccc2)cc1)NCC1CCN(Cc2ccccc2)CC1', 'CCOc1ccc(C(=O)Cc2ncc(C(=O)N(C)CCCOc3ccc([N+](=O)[O-])cc3)cn2)cc1', 'NC(=O)c1cc[n+](CCC[n+]2cc(F)c(/C=N/O)c(F)c2)cc1.[Br-].[Br-]', 'COc1ccc(C(=O)Cc2ncc(C(=O)N(C)CCCOc3ccc(C(F)(F)F)cc3)cn2)cc1', 'COc1ccc(C(=O)Cc2ncc(C(=O)NCc3ccc(OC)c(OC)c3)cn2)cc1', 'Cc1cccc2sc(NC(=O)CCN3CCN(C(C)(C)C)CC3)nc12', 'COc1ccc(/C=C/C(=O)Nc2cc[n+](Cc3cccc(F)c3)cc2)cc1OC.[Br-]', 'Cc1cccc2sc(NC(=O)CCN3CCN(Cc4ccc(C(C)(C)C)cc4)CC3)nc12', 'COc1ccc(C(=O)Cc2ncc(C(=O)N(C)CCCOc3ccc([N+](=O)[O-])cc3)cn2)cc1', 'CCN(C)C(=O)Oc1ccc(C2CC(=O)c3c(O)cc(OC(=O)N(C)CC)cc3O2)cc1', 'COc1cc2nc3c(c(N4CCOCC4)c2cc1OC)CCCC3', 'NC(=O)c1cc[n+](CCCC[n+]2ccc(/C=N/O)c(F)c2)cc1.[Br-].[Br-]', 'Cc1cccc2sc(NC(=O)CCN3CCN(Cc4ccccc4F)CC3)nc12', 'O=C(CN1CCN(Cc2ccccc2)CC1)Nc1n[nH]c2ncccc12', 'O=[N+]([O-])c1ccc2nc3c(c(Nc4[nH]nc5ncccc45)c2c1)CCCC3', 'CCC(c1ccc(O)c(OC)c1)c1ccc(OC)cc1O', 'O=c1c2ccccc2nc2n1C(c1cncc3[nH]c4ccccc4c13)CC2O', 'COc1ccc(OCC(=O)NCCN2CCN(Cc3ccccc3)CC2)cc1', 'Cc1ccc2nc3c(c(NCCCl)c2c1)CCCC3', 'O=C(COc1ccc(Cl)cc1)NCC1CCN(Cc2ccccc2)CC1', 'CC(=O)c1ccc(N2CCN(CCC(=O)Nc3nc4c(C)cccc4s3)CC2)cc1', 'COc1cc(/C=C/c2cc(=O)oc3c2ccc2oc(-c4ccccc4)cc23)cc(OC)c1', 'CN1CCN(CCC(=O)Nc2n[nH]c3ncccc23)CC1', 'Cc1ccc2nc3c(c(NCCN4CCN(C)CC4)c2c1)CCCC3', 'Cc1cccc2sc(NC(=O)CCN3CCN(c4ccccn4)CC3)nc12', 'COc1ccc(OCC(=O)NCC2CCN(Cc3ccccc3)CC2)cc1', 'CC(=O)N1N=C(c2ccc(Br)cc2)CC1c1ccc(-c2ccccc2)cc1', 'NC(=O)c1cc[n+](CCC[n+]2ccc(/C=N/O)c(F)c2)cc1.[Br-].[Br-]', 'COc1cc(/C=C/C(=O)OCCCCCN(C)Cc2cccc(OC(=O)N(C)C)c2)ccc1O', 'C=C(C)[C@]1(O)CC[C@@]2(C)C3=Cc4c(cc(-c5cccnc5)oc4=O)O[C@]3(C)[C@@H](O)C[C@@H]2[C@@]12CCC(=O)OC2', 'Cc1cc(Br)cc(C)c1OCC(=O)NCC1CCN(Cc2ccccc2)CC1', 'CCCCC1OC(=O)c2cc(OCc3ccc(CN4CCCCC4)c(O)c3)ccc21', 'CCN(CC)Cc1ccc(C=C2Cc3ccccc3C2=O)cc1', 'CCCCN(CCCc1c[nH]c2ccccc12)CCC1CCN(Cc2ccccc2)CC1', 'O=C(Nc1ccc(OC(F)(F)F)cc1)NC1CCN(C(=O)CCCCNc2c3c(nc4cc(Cl)ccc24)CCCC3)CC1', 'CCCCCCCNC(=O)Oc1ccc(Cc2nc3cc(C(=O)N(CC)CC)ccc3n2CCC(C)C)cc1', 'CCCCCCCNC(=O)Oc1ccc(Cc2nc3cc(C(=O)N(CC)CC)ccc3n2CCC(C)C)cc1', 'CC(=O)N1N=C(c2ccc(-c3ccccc3)cc2)CC1c1ccccc1[N+](=O)[O-]', 'O=C(COc1ccc([N+](=O)[O-])cc1)NCC1CCN(Cc2ccccc2)CC1', 'C[C@H]1CN2C(=O)[C@]34Cc5c([nH]c6c7c(ccc56)OC(C)(C)CC7=O)C(C)(C)[C@@H]3C[C@@]2(C1)C(=O)N4', 'CCCCC1OC(=O)c2cc(OCc3ccc(CN4CCCC4)c(O)c3)ccc21', 'CN(CCC1CCCCCC1)Cc1c[nH]c2ccccc12', 'CCC1C/C(=C\\c2cc(OC)c(OC)c(OC)c2Br)C(=O)c2ccccc21', 'NS(=O)(=O)c1ccc(/N=C/c2cc(Br)cc(Cl)c2O)cc1', 'CCN(CC)Cc1ccc(/C=C2\\Cc3ccccc3C2=O)cc1', 'CCCCC1OC(=O)c2cc(OCc3ccc(CN4CCOCC4)c(O)c3)ccc21', 'c1ccc2c(c1)CCN1Cc3cc(OCCCN4CCCCC4)ccc3N=C21', 'COc1cc(/C=C/c2cc(=O)oc3c2ccc2oc(-c4cc(OC)cc(OC)c4)cc23)cc(OC)c1', 'CCn1c2ccccc2c2cc(/C=C/c3ccc(NC(=S)NCCN4CCCC4)cc3)ccc21', 'CCCCC1OC(=O)c2cc(OCc3ccc(CN(CC)CC)c(O)c3)ccc21', 'O=C(COc1ccc(CCc2ccccc2)cc1)NCC1CCN(Cc2ccccc2)CC1', 'Cc1ccc2nc3c(c(Nc4[nH]nc5ncccc45)c2c1)CCCC3', 'O=C(CCN1CCN(c2ccccc2)CC1)Nc1n[nH]c2ncccc12', 'Cc1cccc2sc(NC(=O)CCN3CCN(C(C)C)CC3)nc12', 'CCCCN(CCC1CCCCCC1)Cc1cc2ccccc2cc1O', 'CCCCC1OC(=O)c2cc(OCc3ccc(CN4CCN(C)CC4)c(O)c3)ccc21', 'OCCCCCCCCc1ccc(CN2CCCCC2)c(O)c1', 'CCN(CC)Cc1ccc(/C=C2\\Cc3ccc(OC)cc3C2=O)cc1', 'Oc1cc2c(cc1O)-c1cc3ccc(O)c(O)c3c[n+]1CC2.[Br-]', 'NC(=O)c1cc[n+](CCCC[n+]2cc(F)c(/C=N/O)c(F)c2)cc1.[Br-].[Br-]', 'CCCC[N+](C)(CCCc1c[nH]c2ccccc12)CCC1CCCCCC1.[I-]', 'CCN1CCN(c2ccc(-c3nc4ccccc4[nH]3)cc2)CC1', 'Cc1ccc2nc3c(c(N4CCN(c5c6c(nc7ccc(C)cc57)CCCC6)CC4)c2c1)CCCC3', 'CC(=O)N1N=C(c2ccc3c(c2)OCO3)CC1c1ccc(-c2ccccc2)cc1', 'COc1cccc(C(=O)Cc2ncc(C(=O)N(C)CCCOc3ccc([N+](=O)[O-])cc3)cn2)c1', 'CCN(CC)Cc1ccc(C=C2Cc3cc(OC)ccc3C2=O)cc1', 'O=C(COc1ccc(CCc2ccccc2)cc1)NCCN1CCN(Cc2ccccc2)CC1', 'COc1cc(/C=C/c2cc(=O)oc3c2ccc2oc(-c4ccc(Cl)cc4)cc23)cc(OC)c1', 'CC(=O)Nc1ccc2nc3c(c(N4CCCCC4)c2c1)CCCC3', 'O=C(CCN1CCCCC1)Nc1n[nH]c2ncccc12', 'CC(=O)Nc1ccc2nc3c(c(N4CCOCC4)c2c1)CCCC3', 'CCN1CCN(CCC(=O)Nc2nc3c(C)cccc3s2)CC1', 'COc1ccc(C(=O)Cc2ncc(C(=O)NC3CCN(Cc4ccccc4)CC3)cn2)cc1', 'O=C(COc1ccc(Cl)cc1[N+](=O)[O-])NCCN1CCN(Cc2ccccc2)CC1', 'COc1ccc(C(=O)Cc2ncc(C(=O)N(C)CCCOc3cccc(OC)c3)cn2)cc1', 'COc1ccc(C(=O)Cc2ncc(C(=O)N(C)CCCOc3c(F)c(F)c(F)c(F)c3F)cn2)cc1', 'O=c1c2ccccc2nc2n1CC[C@@H]2[C@@H]1CCCN1', 'c1ccc2c(c1)CCCN1Cc3cc(OCCCN4CCCCC4)ccc3N=C21', 'COc1ccc(C(=O)Cc2ncc(C(=O)N3CCN(Cc4ccccc4)CC3)cn2)cc1', 'CCCCN(CCc1c[nH]c2c(OC(=O)C(C)C)cccc12)CCC1CCCCCC1', 'CCN(C)C(=O)Oc1ccc(C2CC(=O)c3c(OC(=O)N(C)CC)cc(OC(=O)N(C)CC)cc3O2)cc1', 'CCN(CC)C(=O)c1ccc2c(c1)nc(Cc1ccc(OC(=O)N3CCC3)cc1)n2CCC(C)C', 'CCN(CC)C(=O)c1ccc2c(c1)nc(Cc1ccc(OC(=O)N3CCC3)cc1)n2CCC(C)C', 'C[N+](C)(C)CCNC(=O)/C=C/c1ccc(O)c(O)c1', 'CN(CCCOc1ccc([N+](=O)[O-])cc1)C(=O)c1cnc(CC(=O)c2cccs2)nc1', 'CC(=O)N1N=C(c2ccc3ccccc3c2)CC1c1ccc(-c2ccccc2)cc1', 'CN1c2ccccc2C(NCCCCCCC(=O)CCCCNc2c3c(nc4ccccc24)CCCC3)c2ccc(Cl)cc2S1(=O)=O', 'COc1cc2c(cc1OC)C(=O)/C(=C/c1cc[n+](Cc3ccc(F)cc3)cc1)OC2.[Br-]', 'CCCCCCCNC(=O)Oc1ccc(Cc2nc3cc(C(=O)N(CC)CC)ccc3n2CCCCN2CCCCC2)cc1', 'CCCCCCCNC(=O)Oc1ccc(Cc2nc3cc(C(=O)N(CC)CC)ccc3n2CCCCN2CCCCC2)cc1', 'CC1=C[C@H]2Cc3nc4cc(Cl)ccc4c(NCCCCC(=O)N4CCC(NC(=O)Nc5ccc(OC(F)(F)F)cc5)CC4)c3[C@@H](C1)C2', 'CC(=O)N1N=C(c2ccc(-c3ccccc3)cc2)CC1c1ccc2c(c1)OCO2', 'COc1ccc2cc3[n+](cc2c1OC)CCc1cc(O)c(O)cc1-3.[Br-]', 'Cc1ccc2nc3c(c(NCCN4CCCCC4)c2c1)CCCC3', 'CCCCCCCNC(=O)Oc1ccc(Cc2nc3cc(C(=O)N4CCCCC4)ccc3n2CCC(C)C)cc1', 'CCCCCCCNC(=O)Oc1ccc(Cc2nc3cc(C(=O)N4CCCCC4)ccc3n2CCC(C)C)cc1', 'CCCCN(CCc1c[nH]c2cc(O)ccc12)CCC1CCCCCC1', 'O=C(COc1cc(F)ccc1[N+](=O)[O-])NCCN1CCN(Cc2ccccc2)CC1', 'CC(=O)N1N=C(c2ccc(C)cc2)CC1c1ccc(-c2ccccc2)cc1', 'CN(C)C1CCC(CCN(CCCc2c[nH]c3ccccc23)CCC2CCCCCC2)CC1', 'CN(C)C1CCC(CCN(CCCc2c[nH]c3ccccc23)CCC2CCCCCC2)CC1', 'CCCCN(CCC1CCCCCC1)Cc1ccc(OC(=O)N(C)C)c2ccccc12', 'Cc1ccc2nc3c(c(NCCO)c2c1)CCCC3', 'COc1cc(/C=C/c2nnc(-c3ccc(C(F)(F)F)cc3)o2)ccc1O', 'Cc1cc(Br)cc(C)c1OCC(=O)NCCN1CCN(Cc2ccccc2)CC1', 'CO[C@@H]1C=C[C@@]23c4cc5c(cc4CN(C[C@@H]2OC(=O)c2ccc([N+](=O)[O-])cc2)C3C1)OCO5', 'O=c1c2ccccc2nc2n1C(C1CN=Cc3[nH]c4ccccc4c31)CC2O', 'CC(=O)N1N=C(c2ccccc2[N+](=O)[O-])CC1c1ccc(-c2ccccc2)cc1', 'COc1ccc2nc3c(c(N)c2c1)C1CC=CC(C3)C1', 'O=C(CN1CCN(c2ccccc2)CC1)Nc1n[nH]c2ncccc12', 'CCCCN(CCC1CCCCCC1)Cc1cn(OC(=O)N(C)C)c2ccccc12', 'OCCCCCCCCc1ccc(CN2CCCC2)c(O)c1', 'Cc1cccc2sc(NC(=O)CCN3CCN(Cc4cc(F)ccc4F)CC3)nc12', 'O=[N+]([O-])c1ccc2nc3c(c(N4CCOCC4)c2c1)CCCC3', 'COc1ccc(C(=O)Cc2ncc(C(=O)N(C)CCCOc3ccc(OC)c(OC)c3)cn2)cc1', 'CCCCC1OC(=O)c2cc(OCc3ccc(CN(C)C)c(O)c3)ccc21', 'O=C(COc1ccc([N+](=O)[O-])cc1)NCCN1CCN(Cc2ccccc2)CC1', 'Cc1cccc2sc(NC(=O)CCN3CCN(CCCN(C)C)CC3)nc12', 'O=C(Cn1nc2ncccc2c1NC(=O)Cn1nc2ncccc2c1NC(=O)CN1CCN(Cc2ccccc2)CC1)Nc1[nH]nc2ncccc12', 'O=C(CCN1CCOCC1)Nc1n[nH]c2ncccc12', 'CC(=O)O[C@@H]1C[C@@]2(C)Oc3cc(C)oc(=O)c3C[C@@H]2[C@@]2(C)C=CC(=O)C(C)(C)[C@H]12', 'CC(=O)O[C@@H]1C[C@@]2(C)Oc3cc(C)oc(=O)c3C[C@@H]2[C@@]2(C)C=CC(=O)C(C)(C)[C@H]12', 'NS(=O)(=O)c1ccc(N2N=C(c3cccs3)CC2c2ccc(Br)cc2)cc1', 'O=C(COc1cc(F)ccc1[N+](=O)[O-])NCC1CCN(Cc2ccccc2)CC1', 'O=C(CCN1CCN(Cc2ccccc2)CC1)Nc1ccc2nc3n(c(=O)c2c1)CCC3', 'COc1cc(/C=C/c2cc(=O)oc3c2ccc2oc(-c4ccc(F)cc4)cc23)cc(OC)c1', 'COc1ccc(OCCCN(C)C(=O)c2cnc(CC(=O)c3ccc(OC)cc3)nc2)cc1', 'NS(=O)(=O)c1ccc(NCc2cccc(Br)c2O)cc1', 'Cc1ccc2nc3c(c(N4CCCCC4)c2c1)CCCC3', 'COc1cc(/C=C/c2cc(=O)oc3c2ccc2oc(-c4cccc(Cl)c4)cc23)cc(OC)c1', 'COc1cc(/C=C/c2nnc(-c3ccc(OC(F)(F)F)cc3)o2)ccc1O', 'CC(=O)N1N=C(c2ccc(-c3ccccc3)cc2)CC1c1ccc2ccccc2c1', 'COC/C(C=O)=C\\CCC1=CC(c2cc(O)ccc2O)OC1=O', 'Cc1cccc2sc(NC(=O)CCN3CCN(Cc4cccc5ccccc45)CC3)nc12', 'Cc1ccc2nc3c(c(NCCN4CCN(Cc5ccccc5)CC4)c2c1)CCCC3', 'Cc1cccc2sc(NC(=O)CCN3CCN(C4CCN(C)CC4)CC3)nc12', 'COC(=O)/C=C/c1ccc(OCCCCCN2CCN(CC(=O)Nc3c4c(nc5ccccc35)CCCC4)CC2)c(OC)c1', 'Cc1ccc2nc3c(c(NCCN4CCOCC4)c2c1)CCCC3', 'CN1CCN(c2c3c(nc4ccc([N+](=O)[O-])cc24)CCCC3)CC1', 'COc1ccc(C(=O)Cc2ncc(C(=O)Nc3ccncc3)cn2)cc1', 'O=C(CCN1CCN(Cc2ccccc2)CC1)Nc1n[nH]c2ncccc12', 'C#CCOc1cccc(-c2cc(-c3ccc(OCCN4CCOCC4)cc3)nc(C)n2)c1', 'COc1ccc(-c2cc3c(ccc4c(/C=C/c5cc(OC)cc(OC)c5)cc(=O)oc43)o2)cc1', 'O=C(Nc1ccc(OC(F)(F)F)cc1)NC1CCN(C(=O)CCCNc2c3c(nc4cc(Cl)ccc24)CCCC3)CC1', 'O=c1cc(/N=N/N=P(CCP(=N/N=N/c2cc(=O)oc3ccccc23)(c2ccccc2)c2ccccc2)(c2ccccc2)c2ccccc2)c2ccccc2o1', 'CCOC1(c2cc(O)ccc2O)C=C(CC/C=C(\\C)CCC=C(C)C)C(=O)O1', 'c1cc2c(cc1OCCCN1CCCCC1)CN1CCCCCC1=N2', 'CCCCCCCNC(=O)Oc1ccc(Cc2nc3cc(C(=O)N(CC)CC)ccc3n2CCN2CCCCC2)cc1', 'CCCCCCCNC(=O)Oc1ccc(Cc2nc3cc(C(=O)N(CC)CC)ccc3n2CCN2CCCCC2)cc1', 'COc1cc2oc(C(=O)Nc3cccc(CN(C)Cc4ccccc4)c3)cc(=O)c2cc1OC', 'O=C(COc1ccccc1C(F)(F)F)NCC1CCN(Cc2ccccc2)CC1', 'COc1cc(/C=C2\\CCc3cc([N+](=O)[O-])ccc3C2=O)c(Br)c(OC)c1OC', 'COc1cc(/C=C/c2cc(=O)oc3c2ccc2oc(-c4cccc(F)c4)cc23)cc(OC)c1', 'O=[N+]([O-])c1ccc2nc3c(c(N4CCCCC4)c2c1)CCCC3', 'Cc1ccc2nc3c(c(N4CCN(C)CC4)c2c1)CCCC3', 'CCOCCN1CCN(CCC(=O)Nc2nc3c(C)cccc3s2)CC1', 'O=C(COc1ccc(Cc2ccccc2)cc1)NCC1CCN(Cc2ccccc2)CC1', 'Cc1cccc2sc(NC(=O)CCN3CCN(C4CCCCC4)CC3)nc12', 'COc1ccc(C(=O)Cc2ncc(C(=O)N(C)CCCOc3ccc([N+](=O)[O-])cc3)cn2)cc1OC', 'CCN(CC)Cc1ccc(C=C2Cc3ccc(OC)cc3C2=O)cc1', 'O=C(COc1ccc(-c2ccccc2)cc1)NCCN1CCN(Cc2ccccc2)CC1', 'CN(CCCOc1ccc([N+](=O)[O-])cc1)C(=O)c1cnc(CC(=O)c2ccccc2)nc1', 'O=C(Nc1ccc(OC(F)(F)F)cc1)NC1CCN(C(=O)CCNc2c3c(nc4cc(Cl)ccc24)CCCC3)CC1', 'Cc1cccc2sc(NC(=O)CCN3CCN(CC4CCCCC4)CC3)nc12', 'Cc1cccc2sc(NC(=O)CCN3CCN(CCN(C(C)C)C(C)C)CC3)nc12', 'COc1ccc(C(=O)Cc2ncc(C(=O)NCc3cccs3)cn2)cc1', 'Nc1nn(-c2c3c(nc4ccc([N+](=O)[O-])cc24)CCCC3)c2ncccc12', 'COc1ccc(/C=C2\\CCc3cc([N+](=O)[O-])ccc3C2=O)c(Cl)c1OC', 'CCN(CC)Cc1ccc(C=C2Cc3cc(OC)c(OC)cc3C2=O)cc1', 'O=C(COc1ccc(Cc2ccccc2)cc1)NCCN1CCN(Cc2ccccc2)CC1', 'Cc1cccc2sc(NC(=O)CCN3CCN(c4ccncc4)CC3)nc12', 'COc1ccc(/C=C/C(=O)Nc2cc[n+](Cc3ccc(F)cc3)cc2)cc1OC.[Br-]', 'NC(=O)c1cc[n+](C/C=C/C[n+]2cc(F)c(/C=N/O)c(F)c2)cc1.[Br-].[Br-]', 'Cc1cccc2sc(NC(=O)CCN3CCN(c4ncccn4)CC3)nc12', 'O=C(COc1ccc(Cl)cc1[N+](=O)[O-])NCC1CCN(Cc2ccccc2)CC1', 'CN(CCC1CCN(Cc2ccccc2)CC1)Cc1c[nH]c2ccccc12', 'CCN(CC)Cc1ccc(/C=C2\\Cc3cc(OC)ccc3C2=O)cc1', 'CC1=C[C@@H]2Cc3nc4cc(Cl)ccc4c(NCCCCC(=O)N4CCC(NC(=O)Nc5ccc(OC(F)(F)F)cc5)CC4)c3[C@H](C1)C2', 'Cc1cccc2sc(NC(=O)CCN3CCN(CC4CCCO4)CC3)nc12', 'Cc1cccc2sc(NC(=O)CCN3CCN(CCN(C)C)CC3)nc12', 'COc1ccccc1/C=C/C=C(\\Br)S(=O)(=O)F', 'CN(CCCOc1ccc([N+](=O)[O-])cc1)C(=O)c1cnc(CC(=O)c2ccc3c(c2)OCO3)nc1', 'c1ccc2c(c1)CN1Cc3cc(OCCCN4CCCCC4)ccc3N=C21', 'c1cc2c(cc1OCCCN1CCCCC1)CN1CCCCC1=N2', 'COc1cc2c(Nc3ccc(C(=O)/C=C/c4nc(C)c(C)nc4C)cc3)ncnc2c(OC)c1OC', 'O=C(COc1ccccc1C(F)(F)F)NCCN1CCN(Cc2ccccc2)CC1', 'Cc1ccc2nc3c(c(N4CCOCC4)c2c1)CCCC3', 'CCN(CC)Cc1ccc(/C=C2/Cc3cc(OC)ccc3C2=O)cc1', 'CN(CCCOc1ccc2ccc(=O)oc2c1)Cc1cccc(OC(=O)NC2CCCC2)c1', 'O=C(Oc1cc(O)c2c(=O)cc(-c3ccc(O)cc3)oc2c1)N1CCCCC1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2ccc(-c3cc(=O)c4ccccc4o3)cc2)c1', 'COc1cccc2c1CCC(NC(=O)OCc1ccccc1)C2', 'CNC(=O)Oc1cccc(CN(C)CCCCCCOc2ccc(C(=O)/C=C/c3cc(OC)c(OC)c(OC)c3)c(O)c2)c1', 'CC(=O)N1CC[C@@]2(O)c3cc(OC(=O)Nc4ccccc4)ccc3N[C@@H]12', 'CC[C@@]1(c2cccc(OC(=O)Nc3ccccc3)c2)CCCCN(C)C1', 'O=c1c(NCCCCCCCc2c3c(nc4ccccc24)CCCC3)c(NCCCCCCCc2c3c(nc4ccccc24)CCCC3)c1=O', 'COc1cc2c(c(OC)c1)CC(NC(=O)OCc1ccccc1)C2', 'CC1CCCCN1CCCNC(=O)c1cc(NC(=O)OC(C)(C)C)ccc1O', 'CCN1/C(=C/C(C)=O)Sc2ccc(OC(=O)OCCOCCNC(=O)c3cc4c(OC(=O)N(C)C)cccc4[n+](C)c3)cc21.COS(=O)(=O)[O-]', 'CCOC(=O)c1cc2c(OC(=O)N(C)C)cccc2[n+](C)c1.O=C([O-])C(F)(F)F', 'CCN(CC)C(=O)OC1C[N+]2(C)CCC1CC2.[I-]', 'CNC(=O)Oc1cccc2c1CCC1C2CCN1CCC1CCCCC1', 'Cc1c(Cl)c(=O)oc2cc(OCCCCSC(=S)N3C(C)CCCC3C)ccc12', 'Cc1c(Cl)c(=O)oc2cc(OCCCCSC(=S)N3CCCCC3C)ccc12', 'O=c1[nH]c2ccc(OCc3ccc(F)cc3)cc2c(=O)o1', 'CNC(=O)Oc1cccc(CN(C)CCCCCCCOc2cccc(-c3cc(=O)c4ccccc4o3)c2)c1', 'COc1cccc2c1CCCC2NS(=O)(=O)NC(=O)OCc1ccccc1', 'CC(C)c1ccc(COc2ccc3[nH]c(=O)oc(=O)c3c2)cc1', 'CCN(C)C(=O)Oc1cccc(CN2CCC(CCC(=O)c3cc(Cl)c(N)cc3OC)CC2)c1', 'O=c1c(NCCCCc2c3c(nc4cc(Cl)ccc24)CCCC3)c(NCCCCc2c3c(nc4ccc(Cl)cc24)CCCC3)c1=O', 'CNC(=O)Oc1cccc(CN(C)CCCCCCOc2ccc3c(c2)O/C(=C\\c2cc(OC)c(OC)c(OC)c2)C3=O)c1', 'CN(CCCOc1ccc2ccc(=O)oc2c1)Cc1cccc(OC(=O)NC2CCCCC2)c1', 'CCC(C)(C)NCC(O)c1cc(O)cc(OC(=O)N(C)C)c1', 'CC(C)(C)OC(=O)Nc1ccc(O)c(C(=O)NCCCN2CCCCC2)c1', 'O=C1CCOc2cc(OCCCCCSC(=S)N3CCCCC3)ccc21', 'Cn1ccc2cc(-c3cnc4ccc(C(=O)N5CCCCC5)cc4n3)ccc2c1=O', 'COC(=O)c1cc2c(OC(=O)N(C)C)cccc2[n+]([11CH3])c1']

Also, manually checking for any “NaNs” in the canonical SMILES column (X variable), since AdaBoost classifier won’t accept missing values in the dataset, but if using HistGradientBoostingClassifier() instead, it should take care of the native NaNs.

print(f"{df_ache.canonical_smiles.isna().sum()} out of {len(df_ache)} SMILES failed in conversion")
0 out of 7075 SMILES failed in conversion

There are other ways to deal with NaNs with a few examples provided by scikit-learn. However, with regards to drug discovery data, there are probably more caveats that need to be taken during data preprocessing (I’m also still exploring this too).


Split data

Randomly splitting data this time.

# Found a silly error when naming X, y train/test sets!
# Remember to name them in this order: X_train, X_test, y_train, y_test
# otherwise model fitting won't work...
X_train, X_test, y_train, y_test = train_test_split(X_valid, y, test_size=0.2, random_state=3)


Create pipelines

The aim is to create pipeline(s) using scikit-learn.


AdaBoost classifier

The original plan is to chain an AdaBoost classifier, XGBoost classifier, along with Scikit-mol transformers all at once. However, it turns out that I’m building two separate pipelines of AdaBoost classifier and XGBoost classifier so that I can compare the difference(s) between them, and this also serves better for the purpose of this post really.

This is also the time to think about generating molecular features for model training. Choosing data features such as fingerprints (e.g. Morgan fingerprints which is usually best for larger dataset) or RDKit 2D descriptors (which is useful for smaller datasets) or others. For RDKit 2D descriptors, Scikit-mol has integrated RDKit’s rdkit.Chem.Descriptors module and rdkit.ML.Descriptors.MoleculeDescriptors module within its MolecularDescriptorTransformer().

Some useful links regarding building pipelines in scikit-learn and also another reference notebook on when to use parallel calculations for different molecular features:

  • Pipeline module

  • make_pipeline module (simpler pipeline construction, without naming estimators ourselves but rather naming them automatically)

  • Useful notebook explaining when is the best to run parallel calculations to calculate molecular fingerprints and descriptors

Initally, RDKit 2D descriptors are used and to see all the available descriptors, run the following code.

rdkit2d = MolecularDescriptorTransformer()
available_descriptors = rdkit2d.available_descriptors
print(f"First 10 descriptor names: {available_descriptors[:10]}")
First 10 descriptor names: ['MaxAbsEStateIndex', 'MaxEStateIndex', 'MinAbsEStateIndex', 'MinEStateIndex', 'qed', 'SPS', 'MolWt', 'HeavyAtomMolWt', 'ExactMolWt', 'NumValenceElectrons']

For the first sample pipeline I’m building below, I’ve noticed that not all of the 209 RDKit 2D descriptors can be used for AdaBoost classifier, as some of the descriptors will have values of “0” and AdaBoost classifier will not be able to take care of them. Therefore, I’m only using a small selection of descriptors only, but HistGradientBoostingClassifier() should be able to take into account of NaNs and can be chained to include all descriptors in a pipeline.

The following is an example of building a scikit_learn pipeline by using AdaBoost classifier model, along with Scikit-mol’s transformers for multi-class max_phase predictions with training set consisting of molecules with max_phase 0, 1, 2, 3 and 4. I’ve used Morgan fingerprints instead eventually so that’ll be shown in the following pipeline code, but I’ve also kept the RDKit 2D descriptor option on (just need to uncomment to run).

# Set parameters for RDKit 2D descriptors
# params_rdkit2d = {
#     "desc_list": ['HeavyAtomCount', 'FractionCSP3', 'RingCount', 'MolLogP', 'MolWt']
# }

# Set parameters for adaboost model
params_adaboost = {
    "estimator": DecisionTreeClassifier(max_depth = 3), 
    # default: n_estimators = 50, learning_rate = 1.0 (trade-off between them)
    "n_estimators": 80, 
    "learning_rate": 0.2, 
    # SAMME (Stagewise Additive Modeling using a Multi-class Exponential loss function) algorithm 
    # for multi-class classification
    "algorithm": "SAMME", 
    "random_state": 2,
    }

# Building AdaBoostClassifier pipeline
mlpipe_adaboost = make_pipeline(
    # Convert SMILES to RDKit molecules
    SmilesToMolTransformer(), 
    # Molecule standardisations
    Standardizer(),
    ## A choice of using either Morgan fingerprints or RDKit 2D descriptors:
    # Generate MorganFingerprintTransformer()
    MorganFingerprintTransformer(useFeatures=True),
    # Generate RDKit2D descriptors
    #MolecularDescriptorTransformer(**params_rdkit2d),
    # Scale variances in descriptor data
    StandardScaler(),
    # Apply adaptive boost classifier
    AdaBoostClassifier(**params_adaboost)
)
# Check on pipeline
mlpipe_adaboost
Pipeline(steps=[('smilestomoltransformer', SmilesToMolTransformer()),
                ('standardizer', Standardizer()),
                ('morganfingerprinttransformer',
                 MorganFingerprintTransformer(useFeatures=True)),
                ('standardscaler', StandardScaler()),
                ('adaboostclassifier',
                 AdaBoostClassifier(algorithm='SAMME',
                                    estimator=DecisionTreeClassifier(max_depth=3),
                                    learning_rate=0.2, n_estimators=80,
                                    random_state=2))])
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

An interactive pipeline diagram should show with a status of “not fitted” if you hover the mouse over the “i” logo on top right. The pipeline is then fitted onto the training sets.

mlpipe_adaboost.fit(X_train, y_train)
Pipeline(steps=[('smilestomoltransformer', SmilesToMolTransformer()),
                ('standardizer', Standardizer()),
                ('morganfingerprinttransformer',
                 MorganFingerprintTransformer(useFeatures=True)),
                ('standardscaler', StandardScaler()),
                ('adaboostclassifier',
                 AdaBoostClassifier(algorithm='SAMME',
                                    estimator=DecisionTreeClassifier(max_depth=3),
                                    learning_rate=0.2, n_estimators=80,
                                    random_state=2))])
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

The pipeline status should now show a “fitted” message if hovering over the same “i” logo. Then the pipeline is used on the X_test (testing X set) to predict the target (max_phase) variable.

mlpipe_adaboost.predict(X_test)
array([0, 0, 0, ..., 0, 0, 0])


XGBoost classfier

The following code snippet is an example of a scikit_learn pipeline using Scikit-mol’s transformers and XGBoost classifier. One nice thing about XGBoost is that is has scikit_learn interface so that it can be integrated into the scikit-learn pipeline and Scikit-mol’s transformers, which is what I’ve tried below.

# Set parameters for xgboost model
params_xgboost = {
    "n_estimators": 100,
    "max_depth": 3,
    # For multi-class classification, use softprob for loss function (learning task parameters)
    # source: https://xgboost.readthedocs.io/en/latest/parameter.html#learning-task-parameters
    "objective": 'multi:softprob', 
    "learning_rate": 0.1, 
    "subsample": 0.5, 
    "random_state": 2
    }

# Building XGBoostClassifier pipeline
mlpipe_xgb = make_pipeline(
    # Convert SMILES to RDKit molecules
    SmilesToMolTransformer(), 
    # Molecule standardisations
    Standardizer(),
    ## A choice of using either Morgan fingerprints  or RDKit 2D descriptors:
    # Generate MorganFingerprintTransformer()
    MorganFingerprintTransformer(useFeatures=True),
    # Generate RDKit2D descriptors
    #MolecularDescriptorTransformer(**params_rdkit2d),
    # Scale variances in descriptor data
    StandardScaler(),
    # XGBoost classifier
    XGBClassifier(**params_xgboost)
)
mlpipe_xgb
Pipeline(steps=[('smilestomoltransformer', SmilesToMolTransformer()),
                ('standardizer', Standardizer()),
                ('morganfingerprinttransformer',
                 MorganFingerprintTransformer(useFeatures=True)),
                ('standardscaler', StandardScaler()),
                ('xgbclassifier',
                 XGBClassifier(base_score=None, booster=None, callbacks=None,
                               colsample_bylevel=None, colsample_bynode=None,
                               colsample_bytr...
                               feature_types=None, gamma=None, grow_policy=None,
                               importance_type=None,
                               interaction_constraints=None, learning_rate=0.1,
                               max_bin=None, max_cat_threshold=None,
                               max_cat_to_onehot=None, max_delta_step=None,
                               max_depth=3, max_leaves=None,
                               min_child_weight=None, missing=nan,
                               monotone_constraints=None, multi_strategy=None,
                               n_estimators=100, n_jobs=None,
                               num_parallel_tree=None,
                               objective='multi:softprob', ...))])
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
mlpipe_xgb.fit(X_train, y_train)
Pipeline(steps=[('smilestomoltransformer', SmilesToMolTransformer()),
                ('standardizer', Standardizer()),
                ('morganfingerprinttransformer',
                 MorganFingerprintTransformer(useFeatures=True)),
                ('standardscaler', StandardScaler()),
                ('xgbclassifier',
                 XGBClassifier(base_score=None, booster=None, callbacks=None,
                               colsample_bylevel=None, colsample_bynode=None,
                               colsample_bytr...
                               feature_types=None, gamma=None, grow_policy=None,
                               importance_type=None,
                               interaction_constraints=None, learning_rate=0.1,
                               max_bin=None, max_cat_threshold=None,
                               max_cat_to_onehot=None, max_delta_step=None,
                               max_depth=3, max_leaves=None,
                               min_child_weight=None, missing=nan,
                               monotone_constraints=None, multi_strategy=None,
                               n_estimators=100, n_jobs=None,
                               num_parallel_tree=None,
                               objective='multi:softprob', ...))])
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
pred = mlpipe_xgb.predict(X_test)
pred
array([0, 0, 0, ..., 0, 0, 0])


Model metrics

One can never just leave the process of building a machine learning model without evaluating it. Although what I have done below is probably very minimal but it’s somewhat a starting point to think about how good the model is.

# from sklearn.metrics import accuracy_score

# # Following misclassification score function code borrowed and adapted from:
# # https://scikit-learn.org/stable/auto_examples/ensemble/plot_adaboost_multiclass.html#sphx-glr-auto-examples-ensemble-plot-adaboost-multiclass-py

# def misclassification_error(y_true, y_pred):
#     return 1 - accuracy_score(y_true, y_pred)

# mlpipe_adaboost_misclassification_error = misclassification_error(
#     y_test, mlpipe_adaboost.fit(X_train, y_train).predict(X_test)
# )

# mlpipe_xgb_misclassifiaction_error = misclassification_error(
#     y_test, mlpipe_xgb.fit(X_train, y_train).predict(X_test)
# )

# print("Training score for mlpipe_adaboost: "f"{mlpipe_adaboost.score(X_train, y_train):0.2f}")
# print("Testing score for mlpipe_adaboost: "f"{mlpipe_adaboost.score(X_test, y_test):0.2f}")
# print("AdaBoostClassifier's misclassification_error: "f"{mlpipe_adaboost_misclassification_error:0.3f}")

# print("Training score for mlpipe_xgb: "f"{mlpipe_xgb.score(X_train, y_train):0.2f}")
# print("Testing score for mlpipe_xgb: "f"{mlpipe_xgb.score(X_test, y_test):0.2f}")
# print("XGBClassifier's missclassification_error: "f"{mlpipe_xgb_misclassifiaction_error:0.3f}")

An example of code output from running above misclassification error code

It appears that XGBoost model obtained a better prediction accuracy than the AdaBoost one (although the models are built in a very simple way, but it still shows the slight difference in performance). The training data being used here is also very imbalanced with a lot of them being max_phase of “0” than “4”, and with max_phase “4” being our ultimate aim, the dataset used above is really for demonstration only. Also, since this post is already quite long, I’d rather not make this post into a gigantic tl;dr, so for the imbalanced data discussion and exploration, my previous posts have tried to touch on this topic - “Random forest” and “Random forest classifier”.


Hyperparameter tuning for XGBoost classifier

For XGBoost, one of the main things is to minimise model overfitting where several parameters will play important roles to achieve this. For example, learning_rate and subsample are the first two mentioned previously, and another technique is based on regularisation which includes two other parameters, reg_alpha (L1 regularisation based on Manhattan distance) and reg_lamda (L2 regularisation based on Euclidean distance). Both of these regularisation parameters aim to penalise XGBoost’s model complexity to make it a bit more conservative in order to reduce overfitting (Bruce, Bruce, and Gedeck 2020).

A full list of XGBoost classifier pipeline parameters and settings used can be retrieved as shown below. It contains a long list of parameters and one of the ways to find the optimal set of parameters is by using cross-validation (CV).

mlpipe_xgb.get_params()
{'memory': None,
 'steps': [('smilestomoltransformer', SmilesToMolTransformer()),
  ('standardizer', Standardizer()),
  ('morganfingerprinttransformer',
   MorganFingerprintTransformer(useFeatures=True)),
  ('standardscaler', StandardScaler()),
  ('xgbclassifier',
   XGBClassifier(base_score=None, booster=None, callbacks=None,
                 colsample_bylevel=None, colsample_bynode=None,
                 colsample_bytree=None, device=None, early_stopping_rounds=None,
                 enable_categorical=False, eval_metric=None, feature_types=None,
                 gamma=None, grow_policy=None, importance_type=None,
                 interaction_constraints=None, learning_rate=0.1, max_bin=None,
                 max_cat_threshold=None, max_cat_to_onehot=None,
                 max_delta_step=None, max_depth=3, max_leaves=None,
                 min_child_weight=None, missing=nan, monotone_constraints=None,
                 multi_strategy=None, n_estimators=100, n_jobs=None,
                 num_parallel_tree=None, objective='multi:softprob', ...))],
 'verbose': False,
 'smilestomoltransformer': SmilesToMolTransformer(),
 'standardizer': Standardizer(),
 'morganfingerprinttransformer': MorganFingerprintTransformer(useFeatures=True),
 'standardscaler': StandardScaler(),
 'xgbclassifier': XGBClassifier(base_score=None, booster=None, callbacks=None,
               colsample_bylevel=None, colsample_bynode=None,
               colsample_bytree=None, device=None, early_stopping_rounds=None,
               enable_categorical=False, eval_metric=None, feature_types=None,
               gamma=None, grow_policy=None, importance_type=None,
               interaction_constraints=None, learning_rate=0.1, max_bin=None,
               max_cat_threshold=None, max_cat_to_onehot=None,
               max_delta_step=None, max_depth=3, max_leaves=None,
               min_child_weight=None, missing=nan, monotone_constraints=None,
               multi_strategy=None, n_estimators=100, n_jobs=None,
               num_parallel_tree=None, objective='multi:softprob', ...),
 'smilestomoltransformer__parallel': False,
 'smilestomoltransformer__safe_inference_mode': False,
 'standardizer__neutralize': True,
 'standardizer__parallel': False,
 'standardizer__safe_inference_mode': False,
 'morganfingerprinttransformer__fpSize': 2048,
 'morganfingerprinttransformer__parallel': False,
 'morganfingerprinttransformer__radius': 2,
 'morganfingerprinttransformer__safe_inference_mode': False,
 'morganfingerprinttransformer__useBondTypes': True,
 'morganfingerprinttransformer__useChirality': False,
 'morganfingerprinttransformer__useCounts': False,
 'morganfingerprinttransformer__useFeatures': True,
 'standardscaler__copy': True,
 'standardscaler__with_mean': True,
 'standardscaler__with_std': True,
 'xgbclassifier__objective': 'multi:softprob',
 'xgbclassifier__base_score': None,
 'xgbclassifier__booster': None,
 'xgbclassifier__callbacks': None,
 'xgbclassifier__colsample_bylevel': None,
 'xgbclassifier__colsample_bynode': None,
 'xgbclassifier__colsample_bytree': None,
 'xgbclassifier__device': None,
 'xgbclassifier__early_stopping_rounds': None,
 'xgbclassifier__enable_categorical': False,
 'xgbclassifier__eval_metric': None,
 'xgbclassifier__feature_types': None,
 'xgbclassifier__gamma': None,
 'xgbclassifier__grow_policy': None,
 'xgbclassifier__importance_type': None,
 'xgbclassifier__interaction_constraints': None,
 'xgbclassifier__learning_rate': 0.1,
 'xgbclassifier__max_bin': None,
 'xgbclassifier__max_cat_threshold': None,
 'xgbclassifier__max_cat_to_onehot': None,
 'xgbclassifier__max_delta_step': None,
 'xgbclassifier__max_depth': 3,
 'xgbclassifier__max_leaves': None,
 'xgbclassifier__min_child_weight': None,
 'xgbclassifier__missing': nan,
 'xgbclassifier__monotone_constraints': None,
 'xgbclassifier__multi_strategy': None,
 'xgbclassifier__n_estimators': 100,
 'xgbclassifier__n_jobs': None,
 'xgbclassifier__num_parallel_tree': None,
 'xgbclassifier__random_state': 2,
 'xgbclassifier__reg_alpha': None,
 'xgbclassifier__reg_lambda': None,
 'xgbclassifier__sampling_method': None,
 'xgbclassifier__scale_pos_weight': None,
 'xgbclassifier__subsample': 0.5,
 'xgbclassifier__tree_method': None,
 'xgbclassifier__validate_parameters': None,
 'xgbclassifier__verbosity': None}
Note

To see the default values or types of each XGBoost parameter, this XGBoost documentation link is useful (which can be cross-referenced with XGBoost’s Python API reference).

# To obtain only the parameter names for the ease of reading
mlpipe_xgb.get_params().keys()
dict_keys(['memory', 'steps', 'verbose', 'smilestomoltransformer', 'standardizer', 'morganfingerprinttransformer', 'standardscaler', 'xgbclassifier', 'smilestomoltransformer__parallel', 'smilestomoltransformer__safe_inference_mode', 'standardizer__neutralize', 'standardizer__parallel', 'standardizer__safe_inference_mode', 'morganfingerprinttransformer__fpSize', 'morganfingerprinttransformer__parallel', 'morganfingerprinttransformer__radius', 'morganfingerprinttransformer__safe_inference_mode', 'morganfingerprinttransformer__useBondTypes', 'morganfingerprinttransformer__useChirality', 'morganfingerprinttransformer__useCounts', 'morganfingerprinttransformer__useFeatures', 'standardscaler__copy', 'standardscaler__with_mean', 'standardscaler__with_std', 'xgbclassifier__objective', 'xgbclassifier__base_score', 'xgbclassifier__booster', 'xgbclassifier__callbacks', 'xgbclassifier__colsample_bylevel', 'xgbclassifier__colsample_bynode', 'xgbclassifier__colsample_bytree', 'xgbclassifier__device', 'xgbclassifier__early_stopping_rounds', 'xgbclassifier__enable_categorical', 'xgbclassifier__eval_metric', 'xgbclassifier__feature_types', 'xgbclassifier__gamma', 'xgbclassifier__grow_policy', 'xgbclassifier__importance_type', 'xgbclassifier__interaction_constraints', 'xgbclassifier__learning_rate', 'xgbclassifier__max_bin', 'xgbclassifier__max_cat_threshold', 'xgbclassifier__max_cat_to_onehot', 'xgbclassifier__max_delta_step', 'xgbclassifier__max_depth', 'xgbclassifier__max_leaves', 'xgbclassifier__min_child_weight', 'xgbclassifier__missing', 'xgbclassifier__monotone_constraints', 'xgbclassifier__multi_strategy', 'xgbclassifier__n_estimators', 'xgbclassifier__n_jobs', 'xgbclassifier__num_parallel_tree', 'xgbclassifier__random_state', 'xgbclassifier__reg_alpha', 'xgbclassifier__reg_lambda', 'xgbclassifier__sampling_method', 'xgbclassifier__scale_pos_weight', 'xgbclassifier__subsample', 'xgbclassifier__tree_method', 'xgbclassifier__validate_parameters', 'xgbclassifier__verbosity'])

Some of the main XGBoost parameters that can be tuned are n_estimators, max_depth, learning_rate, subsample and reg_lambda. Here, I’m going to try to look for the best combination of learning_rate and subsample for a XGBoost classifier model for now.

# Specify parameters and distributions to be sampled
params_dist = {
    # learning_rate usually between 0.01 - 0.1 as suggested by Raschka et al. 
    # default is between 0 and 1
    "xgbclassifier__learning_rate": [0.05, 0.1, 0.3], 
    # subsample default is between 0 and 1
    "xgbclassifier__subsample": [0.5, 0.7, 1.0]
}


Randomised search CV

The following chunk of code is an example of running randomised search CV. I’ve deliberately folded the code to minimise the reading length of the post and also because the result from it is very similar to the grid search CV used below (randomised search CV run time was 13 min 33.2 seconds due to having two pipelines containing two different machine learning models for the same set of data). It’s being kept as a code reference for anyone who’d like to try it and also as an alternative way to do hyperparameter tuning.

Code 
## Uncomment code below to run
# from sklearn.model_selection import RandomizedSearchCV
# from time import time

## Borrowing a utility function code from scikit_learn documentation to report best scores
## Source: https://scikit-learn.org/stable/auto_examples/model_selection/plot_randomized_search.html#sphx-glr-auto-examples-model-selection-plot-randomized-search-py

# def report(results, n_top=3):
#     for i in range(1, n_top + 1):
#         candidates = np.flatnonzero(results["rank_test_score"] == i)
#         for candidate in candidates:
#             print("Model with rank: {0}".format(i))
#             print(
#                 "Mean validation score: {0:.3f} (std: {1:.3f})".format(
#                     results["mean_test_score"][candidate],
#                     results["std_test_score"][candidate],
#                 )
#             )
#             print("Parameters: {0}".format(results["params"][candidate]))
#             print("")

## The following code has also referenced and adapted from this notebook 
## https://github.com/EBjerrum/scikit-mol/blob/main/notebooks/06_hyperparameter_tuning.ipynb

# n_iter_search = 9

# random_search = RandomizedSearchCV(
#     mlpipe_xgb, 
#     param_distributions=params_dist,
#     n_iter=n_iter_search,
#     n_jobs=2
# )

# t_start = time()
# random_search.fit(X_train, y_train)
# t_finish = time()

# print(f'Runtime: {t_finish-t_start:0.2F} seconds for {n_iter_search} iterations')

## Run report function code
# report(random_search.cv_results_)


Grid search CV
# grid_search = GridSearchCV(
#     mlpipe_xgb,
#     param_grid=params_dist,
#     verbose=1,
#     n_jobs=2
# )

# grid_search.fit(X_train, y_train)

# print(f"The best cv score is: {grid_search.best_score_:0.2f}")
# print(f"The best cv parameter settings are: {grid_search.best_params_}")

## This may take longer time to run depending on computer hardware specs (for me it's taken ~13min)

An example of code output from running the grid search code

For tuning parameters of Morgan fingerprints, this Scikit-mol example notebook explains how to do it with code, so I won’t repeat them here, but have only shown how to tune some of the main XGBoost parameters.


Pickle model

The next step is to pickle the model or pipeline if wanting to save it for future use and to avoid re-training model from the ground up again.

Some security tips regarding pickle module:

One thing to remember is to avoid unpickling unknown files over insecure network, and add security key if needed.

# Pickle to save (serialise) the model in working directory (specify path if needed)
# "wb" - write binary
pickle.dump(mlpipe_xgb, open("xgb_pipeline.pkl", "wb"))
# Unpickle (de-serialise) the model
# "rb" - read binary
mlpipe_xgb_2 = pickle.load(open("xgb_pipeline.pkl", "rb"))

# Use the unpickled model object to make prediction
pred2 = mlpipe_xgb_2.predict(X_test)

## Check unpickled model and original model are the same via Python's assertion method
#assert np.sum(np.abs(pred2 - pred)) == 0
## or alternatively use numpy's allclose()
print(np.allclose(pred, pred2))
True


Acknowledgement

Again, this grows into another really long post… Although this post has taken quite a long time to build up to completion, I still want to thank all the contributors or developers for all the packages used in this post.

Back to top

References

Bjerrum, EJ, RA Bachorz, A Bitton, O Choung, Y Chen, C Esposito, SV Ha, and A Poehlmann. 2023. “Scikit-Mol Brings Cheminformatics to Scikit-Learn.” https://chemrxiv.org/engage/chemrxiv/article-details/60ef0fc58825826143a82cc0.
Bruce, Peter, Andrew Bruce, and Peter Gedeck. 2020. Practical Statistics for Data Scientists. https://www.oreilly.com/library/view/practical-statistics-for/9781492072935/.
Chen, Tianqi, and Carlos Guestrin. 2016. “XGBoost: A Scalable Tree Boosting System.” CoRR abs/1603.02754. http://arxiv.org/abs/1603.02754.
Freund, Yoav, and Robert E Schapire. 1997. “A Decision-Theoretic Generalization of On-Line Learning and an Application to Boosting.” Journal of Computer and System Sciences 55 (1): 119–39. https://doi.org/10.1006/jcss.1997.1504.
Raschka, Sebastian, Yuxi (Hayden) Liu, and Vahid Mirjalili. 2022. Machine Learning with PyTorch and Scikit-Learn. Birmingham, UK: Packt Publishing.
Schapire, Robert E. 1990. “The Strength of Weak Learnability.” Machine Learning 5 (2): 197–227. https://doi.org/10.1007/bf00116037.
]]> Machine learning projects Tree models Pandas Scikit-learn RDKit ChEMBL database Python https://jhylin.github.io/Data_in_life_blog/posts/19_ML2-3_Boosted_trees/1_adaboost_xgb.html Wed, 05 Jun 2024 12:00:00 GMT Using Molstar in Quarto Jennifer HY Lin https://jhylin.github.io/Data_in_life_blog/posts/18_Notes_molstar_quarto/Molstar_quarto.html
Background

This is really a short post (note) for myself and probably for others who may be interested in software tools to visualise in silico macromolecules and small molecules.

Most bioinformaticians or structural biologists are probably already familiar with this software package, Molstar or Mol* (Sehnal et al. 2021). Molstar is a 3D viewer for large macromolecules (e.g. proteins), which are commonly used in structural biology and drug discovery (and also other related scientific disciplines).

A Quarto extension has been developed to embed the Molstar interactive 3D viewer inside Quarto markdown documents, which can be rendered as HTML pages. The main advantage of this is that it’s useful for reports or presentations.


An example using Molstar with RCSB PDB

The following example retrieves a protein (PDB ID: 4MQT) from RCSB PDB.

{{< mol-rcsb 4mqt >}}

Hover over protein structure to see details of amino acid residues or ligands present in the structure.

To focus or zoom-in on the ligand bound to the receptor, just click on the ligand first. This shows most of the chemical interactions between the receptor and ligand bound to it (e.g. hydrogen bondings, other chemical interactions will appear if present e.g. pi-pi stacking).

Screenshots or state snapshots are also available from the viewer (other utility functions can be found on the top right corner of the viewer).

An close-up screenshot of 4MQT showing bound ligand

MD trajectories are also available, although I haven’t quite got there yet but it’s useful to know this may be possible (see example C from https://molstar.org/viewer-docs/examples/).

It’s also possible to upload AlphaFold-sourced proteins, or from other file sources (see examples shown from Molstar example).

Back to top

References

Sehnal, David, Sebastian Bittrich, Mandar Deshpande, Radka Svobodová, Karel Berka, Václav Bazgier, Sameer Velankar, Stephen K Burley, Jaroslav Koča, and Alexander S Rose. 2021. “Mol* Viewer: Modern Web App for 3D Visualization and Analysis of Large Biomolecular Structures.” Nucleic Acids Research 49 (W1): W431–37. https://doi.org/10.1093/nar/gkab314.
]]> Notes Molecular viz https://jhylin.github.io/Data_in_life_blog/posts/18_Notes_molstar_quarto/Molstar_quarto.html Fri, 05 Apr 2024 11:00:00 GMT Random forest classifier Jennifer HY Lin https://jhylin.github.io/Data_in_life_blog/posts/17_ML2-2_Random_forest/2_random_forest_classifier.html The section on “Data retrieval using chembl_downloader” has been updated and finalised on 31st January 2024 - many thanks for the comment from Charles Tapley Hoyt (cthoyt).

Jupyter notebook version can be accessed here.


Brief introduction

This post was really just an addition towards the last random forest (RF) post. It was mainly inspired by this paper (Esposito et al. 2021) from rinikerlab1. It was nice to complete the RF series by adding a RF classifier since last post was only on a regressor. Another thing was that imbalanced datasets were common in drug discovery projects, learning different strategies to deal with them was also very useful. While working on this post, I also came across a few other packages that I haven’t used before so I’ve included them all down below.


Overview of post
  • Data sourcing via chembl_downloader
  • Minor data preprocessing using own little script and also SMILES checker from scikit_mol
  • scikit-learn’s RandomForestClassifier()
  • Dealing with imbalanced dataset in RF classifiers by using ghostml
  • A small section on plotting receiver operating characteristic (ROC) curves


Importing libraries
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import chembl_downloader
from chembl_downloader import latest, queries, query
from rdkit.Chem import Descriptors
import datamol as dm
from sklearn.ensemble import RandomForestClassifier
from sklearn import metrics
from sklearn.model_selection import train_test_split
from sklearn.metrics import ConfusionMatrixDisplay
from sklearn.metrics import RocCurveDisplay, roc_curve
from scikit_mol.utilities import CheckSmilesSanitazion
import ghostml


Data retrieval using chembl_downloader

chembl_downloader was something I wanted to try a while back. I’ve tried manual download and chembl_webresource_client previously, and they were probably not the best strategies for data reproducibility. The idea of chembl_downloader was to generate a reproducible ChEMBL data source. It involved some SQL at the beginning to specify the exact type of data needed, so some SQL knowledge were required. Other uses for this package were elaborated much more clearly in its GitHub repository at https://github.com/cthoyt/chembl-downloader. One of the reference notebooks that I’ve used could be reached here (more available in its repository).

What I did was shown below.

# Show the latest version of ChEMBL used
latest_version = latest()
print(f"The latest ChEMBL version is: {latest_version}")
The latest ChEMBL version is: 33

The following section was updated as suggested by cthoyt (via his comment for post below). I ended up putting through my first ever pull request in an open-source and cheminformatics-related repository. A new option to choose max_phase was added into the get_target_sql function in chembl_downloader by keeping it as a togglable option via boolean flag. Many thanks for the patience from cthoyt for guiding me through it. The overall code was now changed as shown below.

# Generate SQL for a query on acetylcholinesterase (AChE): CHEMBL220
sql = queries.get_target_sql(target_id="CHEMBL220", target_type="SINGLE PROTEIN", max_phase=True)

# Pretty-print the SQL in Jupyter
queries.markdown(sql)
SELECT
    ASSAYS.chembl_id              AS assay_chembl_id,
    TARGET_DICTIONARY.target_type,
    TARGET_DICTIONARY.tax_id,
    COMPOUND_STRUCTURES.canonical_smiles,
    MOLECULE_DICTIONARY.chembl_id AS molecule_chembl_id,
    MOLECULE_DICTIONARY.max_phase,
    ACTIVITIES.standard_type,
    ACTIVITIES.pchembl_value
FROM TARGET_DICTIONARY
     JOIN ASSAYS ON TARGET_DICTIONARY.tid == ASSAYS.tid
     JOIN ACTIVITIES ON ASSAYS.assay_id == ACTIVITIES.assay_id
     JOIN MOLECULE_DICTIONARY ON MOLECULE_DICTIONARY.molregno == ACTIVITIES.molregno
     JOIN COMPOUND_STRUCTURES ON MOLECULE_DICTIONARY.molregno == COMPOUND_STRUCTURES.molregno
WHERE TARGET_DICTIONARY.chembl_id = 'CHEMBL220'
    AND ACTIVITIES.pchembl_value IS NOT NULL
    AND TARGET_DICTIONARY.target_type = 'SINGLE PROTEIN'```
:::
:::


::: {.cell execution_count=4}
``` {.python .cell-code}
# The SQL still works as shown above, 
# please ignore the non-SQL texts
# - bit of a formatting issue 
# when pretty-printing the sql, 
# but shouldn't affect the code

I’ve also updated how I retrieved and saved the ChEMBL data with the following code suggested and provided by cthoyt. This would be a better and more reproducible way for anyone who might be interested in re-running this notebook.

from pathlib import Path

# Pick any directory, but make sure it's relative to your home directory
directory = Path.home().joinpath(".data", "blog")
# Create the directory if it doesn't exist
directory.mkdir(exist_ok=True, parents=True)

# Create a file path that corresponds to the version, since this could change
path = directory.joinpath(f"chembl_d_ache_{latest_version}.tsv")

if path.is_file():
    # If the file already exists, load it
    df_ache = pd.read_csv(path, sep=',')
else:
    # If the file doesn't already exist, make the query then cache it
    df_ache = chembl_downloader.query(sql)
    df_ache.to_csv(path, sep=",", index=False)

The rest of the code outputs in the post stayed the same as before. The only thing changed and updated was the part on retrieving ChEMBL data via chembl_downloader.


Some data cleaning

Minor cleaning and preprocessing were done for this post only, as the focus was more on dealing with imbalanced dataset in RF classifier. Since I used a different way to retrieve ChEMBL data this time, the dataset used here might be slightly different from the one used in previous post.


mol_prep.py

I’ve more or less accumulated small pieces of code over time, and I’ve decided to compile them into a Python script. The idea was to remove most function code in the post to avoid repeating them all the time since they’ve been used frequently in the last few posts. The script would be saved into the RF repository, and would still be considered as a “work-in-progress” script (needs more work in the future).

## Trial own mol_prep.py script
from mol_prep import preprocess, rdkit_2d_descriptors
## Preprocess/standardise molecules
# Running preprocess function 
df_ache = df_ache.copy()
df_prep = df_ache.apply(preprocess, axis = 1)
df_prep.head(3)
assay_chembl_id target_type tax_id chembl_id canonical_smiles molecule_chembl_id max_phase standard_type pchembl_value rdkit_mol standard_smiles selfies inchi inchikey
0 CHEMBL1909212 SINGLE PROTEIN 9606 CHEMBL220 CC/C(=C(/CC)c1ccc(O)cc1)c1ccc(O)cc1 CHEMBL411 4.0 IC50 4.59 <rdkit.Chem.rdchem.Mol object at 0x12b7f2180> CC/C(=C(/CC)c1ccc(O)cc1)c1ccc(O)cc1 [C][C][/C][=Branch1][P][=C][Branch1][Ring1][/C... InChI=1S/C18H20O2/c1-3-17(13-5-9-15(19)10-6-13... RGLYKWWBQGJZGM-ISLYRVAYSA-N
1 CHEMBL1003053 SINGLE PROTEIN 9606 CHEMBL220 COc1c2occc2cc2ccc(=O)oc12 CHEMBL416 4.0 IC50 4.27 <rdkit.Chem.rdchem.Mol object at 0x12b7f22d0> COc1c2occc2cc2ccc(=O)oc12 [C][O][C][=C][O][C][=C][C][Ring1][Branch1][=C]... InChI=1S/C12H8O4/c1-14-12-10-8(4-5-15-10)6-7-2... QXKHYNVANLEOEG-UHFFFAOYSA-N
2 CHEMBL2406149 SINGLE PROTEIN 9606 CHEMBL220 COc1c2occc2cc2ccc(=O)oc12 CHEMBL416 4.0 IC50 6.12 <rdkit.Chem.rdchem.Mol object at 0x12b7f2340> COc1c2occc2cc2ccc(=O)oc12 [C][O][C][=C][O][C][=C][C][Ring1][Branch1][=C]... InChI=1S/C12H8O4/c1-14-12-10-8(4-5-15-10)6-7-2... QXKHYNVANLEOEG-UHFFFAOYSA-N


scikit_mol

scikit_mol was a package originated from RDKit UGM hackathon in 2022. This blog post elaborated further on its functions and uses in machine learning. For this post I’ve only used it for a very small portion, mainly to check for missing SMILES or errors in SMILES (kind of like double checking whether the preprocess function code worked as expected). It could be integrated with scikit-learn’s pipeline method for multiple estimators. Its GitHub Repository link: https://github.com/EBjerrum/scikit-mol - I’ve referred to this reference notebook while working on this post.

# Quick simple way to check for missing SMILES
print(f'Dataset contains {df_prep.standard_smiles.isna().sum()} unparsable mols')
Dataset contains 0 unparsable mols
# Checking for invalid SMILES using scikit_mol
smileschecker = CheckSmilesSanitazion()
smileschecker.sanitize(list(df_prep.standard_smiles))

# Showing SMILES errors
smileschecker.errors
SMILES

It showed no errors in SMILES (errors should be listed in the code cell output).

## Generate RDKit 2D descriptors/fingerprints
# Running rdkit_2d_descriptors function
df_2d = rdkit_2d_descriptors(df_prep)
df_2d.head(3)
MaxAbsEStateIndex MaxEStateIndex MinAbsEStateIndex MinEStateIndex qed SPS MolWt HeavyAtomMolWt ExactMolWt NumValenceElectrons ... fr_sulfide fr_sulfonamd fr_sulfone fr_term_acetylene fr_tetrazole fr_thiazole fr_thiocyan fr_thiophene fr_unbrch_alkane fr_urea
0 9.41068 9.41068 0.284153 0.284153 0.779698 12.1000 268.356 248.196 268.146330 104 ... 0 0 0 0 0 0 0 0 0 0
1 11.17310 11.17310 0.405828 -0.405828 0.586359 11.0625 216.192 208.128 216.042259 80 ... 0 0 0 0 0 0 0 0 0 0
2 11.17310 11.17310 0.405828 -0.405828 0.586359 11.0625 216.192 208.128 216.042259 80 ... 0 0 0 0 0 0 0 0 0 0

3 rows × 210 columns

# Merge dataframes df_prep & df_2d via index
df_merge = pd.merge(
    df_prep[["max_phase", "molecule_chembl_id"]],
    df_2d,
    left_index=True,
    right_index=True
)
print(df_merge.shape)
df_merge.head(3)
(7144, 212)
max_phase molecule_chembl_id MaxAbsEStateIndex MaxEStateIndex MinAbsEStateIndex MinEStateIndex qed SPS MolWt HeavyAtomMolWt ... fr_sulfide fr_sulfonamd fr_sulfone fr_term_acetylene fr_tetrazole fr_thiazole fr_thiocyan fr_thiophene fr_unbrch_alkane fr_urea
0 4.0 CHEMBL411 9.41068 9.41068 0.284153 0.284153 0.779698 12.1000 268.356 248.196 ... 0 0 0 0 0 0 0 0 0 0
1 4.0 CHEMBL416 11.17310 11.17310 0.405828 -0.405828 0.586359 11.0625 216.192 208.128 ... 0 0 0 0 0 0 0 0 0 0
2 4.0 CHEMBL416 11.17310 11.17310 0.405828 -0.405828 0.586359 11.0625 216.192 208.128 ... 0 0 0 0 0 0 0 0 0 0

3 rows × 212 columns

A different spreads of max phases were shown this time in the dataframe, as the SQL query mainly used IC50, whereas last post was strictly limited to Ki via ChEMBL web resource client. Other likely reason was that in the decision tree series, I’ve attempted data preprocessing at a larger scale so some data were eliminated. It appeared that there were more max phase 4 compounds here than last time (note: null compounds were not shown in the value counts as it was labelled as “NaN”, it should be the largest max phase portion in the data).

# Find out counts of each max phase
df_merge.value_counts("max_phase")
max_phase
 4.0    618
-1.0     69
 2.0     29
 3.0     10
 1.0      7
 0.5      1
Name: count, dtype: int64

I then tried searching for the chembl_id of the 10 max phase 4 compounds used in the last post in the dataframe (df_merge).

# Previously used 10 max phase 4 compounds
# donepezil = CHEMBL502 & galantamine = CHEMBL659
list_mp4 = ["CHEMBL95", "CHEMBL1128", "CHEMBL640", "CHEMBL502", "CHEMBL481", "CHEMBL360055", "CHEMBL1025", "CHEMBL659", "CHEMBL1200970", "CHEMBL1677"]

# Search for compounds in list_mp4 within df_merge's "molecule_chembl_id" column
# using Series.isin
df_prev = df_merge.loc[df_merge["molecule_chembl_id"].isin(list_mp4)]
df_prev.head()
max_phase molecule_chembl_id MaxAbsEStateIndex MaxEStateIndex MinAbsEStateIndex MinEStateIndex qed SPS MolWt HeavyAtomMolWt ... fr_sulfide fr_sulfonamd fr_sulfone fr_term_acetylene fr_tetrazole fr_thiazole fr_thiocyan fr_thiophene fr_unbrch_alkane fr_urea
9 4.0 CHEMBL481 13.581173 13.581173 0.095133 -1.863974 0.355956 22.209302 586.689 548.385 ... 0 0 0 0 0 0 0 0 0 0
10 4.0 CHEMBL481 13.581173 13.581173 0.095133 -1.863974 0.355956 22.209302 586.689 548.385 ... 0 0 0 0 0 0 0 0 0 0
67 4.0 CHEMBL95 6.199769 6.199769 0.953981 0.953981 0.706488 15.200000 198.269 184.157 ... 0 0 0 0 0 0 0 0 0 0
68 4.0 CHEMBL95 6.199769 6.199769 0.953981 0.953981 0.706488 15.200000 198.269 184.157 ... 0 0 0 0 0 0 0 0 0 0
69 4.0 CHEMBL95 6.199769 6.199769 0.953981 0.953981 0.706488 15.200000 198.269 184.157 ... 0 0 0 0 0 0 0 0 0 0

5 rows × 212 columns

There were many duplicates of compounds.

print(df_prev.shape)
df_prev.value_counts("molecule_chembl_id")
(439, 212)
molecule_chembl_id
CHEMBL95         182
CHEMBL502        143
CHEMBL659         76
CHEMBL1128        12
CHEMBL1677        10
CHEMBL1200970      6
CHEMBL640          4
CHEMBL1025         3
CHEMBL481          2
CHEMBL360055       1
Name: count, dtype: int64
# Dropping duplicated compound via chembl IDs in the main df
df_merge_new = df_merge.drop_duplicates(subset=["molecule_chembl_id"], keep="first")
print(df_merge_new.shape)
df_merge_new.head()
(5357, 212)
max_phase molecule_chembl_id MaxAbsEStateIndex MaxEStateIndex MinAbsEStateIndex MinEStateIndex qed SPS MolWt HeavyAtomMolWt ... fr_sulfide fr_sulfonamd fr_sulfone fr_term_acetylene fr_tetrazole fr_thiazole fr_thiocyan fr_thiophene fr_unbrch_alkane fr_urea
0 4.0 CHEMBL411 9.410680 9.410680 0.284153 0.284153 0.779698 12.100000 268.356 248.196 ... 0 0 0 0 0 0 0 0 0 0
1 4.0 CHEMBL416 11.173100 11.173100 0.405828 -0.405828 0.586359 11.062500 216.192 208.128 ... 0 0 0 0 0 0 0 0 0 0
3 -1.0 CHEMBL7002 11.591481 11.591481 0.189306 -0.309798 0.886859 23.608696 333.453 310.269 ... 1 0 0 0 0 0 0 0 0 0
4 NaN CHEMBL28 12.020910 12.020910 0.018823 -0.410347 0.631833 10.800000 270.240 260.160 ... 0 0 0 0 0 0 0 0 0 0
5 4.0 CHEMBL41 12.564531 12.564531 0.203346 -4.329869 0.851796 12.909091 309.331 291.187 ... 0 0 0 0 0 0 0 0 0 0

5 rows × 212 columns

# Making sure previously used 10 max phase 4 compounds could be found in df_merge_new
df_mp4 = df_merge_new.loc[df_merge_new["molecule_chembl_id"].isin(list_mp4)]
df_mp4
max_phase molecule_chembl_id MaxAbsEStateIndex MaxEStateIndex MinAbsEStateIndex MinEStateIndex qed SPS MolWt HeavyAtomMolWt ... fr_sulfide fr_sulfonamd fr_sulfone fr_term_acetylene fr_tetrazole fr_thiazole fr_thiocyan fr_thiophene fr_unbrch_alkane fr_urea
9 4.0 CHEMBL481 13.581173 13.581173 0.095133 -1.863974 0.355956 22.209302 586.689 548.385 ... 0 0 0 0 0 0 0 0 0 0
67 4.0 CHEMBL95 6.199769 6.199769 0.953981 0.953981 0.706488 15.200000 198.269 184.157 ... 0 0 0 0 0 0 0 0 0 0
249 4.0 CHEMBL502 12.936933 12.936933 0.108783 0.108783 0.747461 20.214286 379.500 350.268 ... 0 0 0 0 0 0 0 0 0 0
508 4.0 CHEMBL640 11.743677 11.743677 0.044300 -0.044300 0.731540 10.529412 235.331 214.163 ... 0 0 0 0 0 0 0 0 0 0
512 4.0 CHEMBL659 9.972866 9.972866 0.008380 -0.411699 0.800524 33.857143 287.359 266.191 ... 0 0 0 0 0 0 0 0 0 0
1013 4.0 CHEMBL1025 12.703056 12.703056 0.426312 -4.304784 0.629869 13.000000 184.147 170.035 ... 0 0 0 0 0 0 0 0 0 0
1345 4.0 CHEMBL1128 9.261910 9.261910 0.000000 0.000000 0.608112 10.692308 201.697 185.569 ... 0 0 0 0 0 0 0 0 0 0
2028 4.0 CHEMBL360055 6.476818 6.476818 0.656759 0.656759 0.205822 12.583333 510.828 450.348 ... 0 0 0 0 0 0 0 0 0 0
2725 4.0 CHEMBL1677 6.199769 6.199769 0.000000 0.000000 0.760853 14.250000 234.730 219.610 ... 0 0 0 0 0 0 0 0 0 0
3271 4.0 CHEMBL1200970 2.520809 2.520809 0.000000 0.000000 0.709785 14.000000 348.943 323.743 ... 0 0 0 0 0 0 0 0 0 0

10 rows × 212 columns

# note: compounds with max phase 0 not shown in the count
df_merge_new.value_counts("max_phase")
max_phase
 4.0    55
-1.0    26
 2.0     8
 3.0     7
 1.0     4
 0.5     1
Name: count, dtype: int64


Model building

The aim of this post was to model and classify the max phases of ChEMBL small molecules, i.e. whether the compounds in the testing set (consisted of max phase 0 or null compounds) might be eventually classified as max phase 4 or not. This was one of the approaches to answer the question in mind, and not the ultimate way to solve the problem (just thought to mention). The target was “max_phase” and features to be used were the various RDKit 2D descriptors (RDKit2D).

The steps I’ve taken to build the model were shown below:

  1. Re-labelled max phases as binary labels (e.g. max phase null as 0, max phase 4 as 1)
# Re-label max phase NaN as 0
df_merge_new = df_merge_new.fillna(0)
df_merge_new.head()
max_phase molecule_chembl_id MaxAbsEStateIndex MaxEStateIndex MinAbsEStateIndex MinEStateIndex qed SPS MolWt HeavyAtomMolWt ... fr_sulfide fr_sulfonamd fr_sulfone fr_term_acetylene fr_tetrazole fr_thiazole fr_thiocyan fr_thiophene fr_unbrch_alkane fr_urea
0 4.0 CHEMBL411 9.410680 9.410680 0.284153 0.284153 0.779698 12.100000 268.356 248.196 ... 0 0 0 0 0 0 0 0 0 0
1 4.0 CHEMBL416 11.173100 11.173100 0.405828 -0.405828 0.586359 11.062500 216.192 208.128 ... 0 0 0 0 0 0 0 0 0 0
3 -1.0 CHEMBL7002 11.591481 11.591481 0.189306 -0.309798 0.886859 23.608696 333.453 310.269 ... 1 0 0 0 0 0 0 0 0 0
4 0.0 CHEMBL28 12.020910 12.020910 0.018823 -0.410347 0.631833 10.800000 270.240 260.160 ... 0 0 0 0 0 0 0 0 0 0
5 4.0 CHEMBL41 12.564531 12.564531 0.203346 -4.329869 0.851796 12.909091 309.331 291.187 ... 0 0 0 0 0 0 0 0 0 0

5 rows × 212 columns

  1. Splitted data into max phase null & max phase 4 (needing to re-label max phase 4 column only as 1, and not disrupting the labels of max phase 0 compounds)
# Select all max phase null compounds
df_null = df_merge_new[df_merge_new["max_phase"] == 0]
print(df_null.shape)
df_null.head()
(5256, 212)
max_phase molecule_chembl_id MaxAbsEStateIndex MaxEStateIndex MinAbsEStateIndex MinEStateIndex qed SPS MolWt HeavyAtomMolWt ... fr_sulfide fr_sulfonamd fr_sulfone fr_term_acetylene fr_tetrazole fr_thiazole fr_thiocyan fr_thiophene fr_unbrch_alkane fr_urea
4 0.0 CHEMBL28 12.020910 12.020910 0.018823 -0.410347 0.631833 10.800000 270.240 260.160 ... 0 0 0 0 0 0 0 0 0 0
8 0.0 CHEMBL8320 10.282778 10.282778 0.120741 -0.120741 0.416681 17.500000 108.096 104.064 ... 0 0 0 0 0 0 0 0 0 0
14 0.0 CHEMBL11833 11.201531 11.201531 0.428520 -0.466092 0.838024 25.157895 262.309 244.165 ... 0 0 0 0 0 0 0 0 0 0
392 0.0 CHEMBL12324 11.257704 11.257704 0.462395 -0.462395 0.797990 26.150000 277.344 256.176 ... 0 0 0 0 0 0 0 0 0 0
393 0.0 CHEMBL274107 11.359778 11.359778 0.372211 -0.473241 0.838024 25.157895 262.309 244.165 ... 0 0 0 0 0 0 0 0 0 0

5 rows × 212 columns

# Using pd.DataFrame.assign to add a new column to re-label max_phase 4 as "1"
df_mp4_lb = df_mp4.assign(max_phase_lb = df_mp4["max_phase"] / 4)

# Using pd.DataFrame.pop() & insert() to shift added column to first column position
first_col = df_mp4_lb.pop("max_phase_lb")
df_mp4_lb.insert(0, "max_phase_lb", first_col)
df_mp4_lb.head()
max_phase_lb max_phase molecule_chembl_id MaxAbsEStateIndex MaxEStateIndex MinAbsEStateIndex MinEStateIndex qed SPS MolWt ... fr_sulfide fr_sulfonamd fr_sulfone fr_term_acetylene fr_tetrazole fr_thiazole fr_thiocyan fr_thiophene fr_unbrch_alkane fr_urea
9 1.0 4.0 CHEMBL481 13.581173 13.581173 0.095133 -1.863974 0.355956 22.209302 586.689 ... 0 0 0 0 0 0 0 0 0 0
67 1.0 4.0 CHEMBL95 6.199769 6.199769 0.953981 0.953981 0.706488 15.200000 198.269 ... 0 0 0 0 0 0 0 0 0 0
249 1.0 4.0 CHEMBL502 12.936933 12.936933 0.108783 0.108783 0.747461 20.214286 379.500 ... 0 0 0 0 0 0 0 0 0 0
508 1.0 4.0 CHEMBL640 11.743677 11.743677 0.044300 -0.044300 0.731540 10.529412 235.331 ... 0 0 0 0 0 0 0 0 0 0
512 1.0 4.0 CHEMBL659 9.972866 9.972866 0.008380 -0.411699 0.800524 33.857143 287.359 ... 0 0 0 0 0 0 0 0 0 0

5 rows × 213 columns

# Also create a new column max_phase_lb column for df_null 
# in order to merge 2 dfs later
df_null_lb = df_null.assign(max_phase_lb = df_null["max_phase"])
first_col_null = df_null_lb.pop("max_phase_lb")
df_null_lb.insert(0, "max_phase_lb", first_col_null)
df_null_lb.head()
max_phase_lb max_phase molecule_chembl_id MaxAbsEStateIndex MaxEStateIndex MinAbsEStateIndex MinEStateIndex qed SPS MolWt ... fr_sulfide fr_sulfonamd fr_sulfone fr_term_acetylene fr_tetrazole fr_thiazole fr_thiocyan fr_thiophene fr_unbrch_alkane fr_urea
4 0.0 0.0 CHEMBL28 12.020910 12.020910 0.018823 -0.410347 0.631833 10.800000 270.240 ... 0 0 0 0 0 0 0 0 0 0
8 0.0 0.0 CHEMBL8320 10.282778 10.282778 0.120741 -0.120741 0.416681 17.500000 108.096 ... 0 0 0 0 0 0 0 0 0 0
14 0.0 0.0 CHEMBL11833 11.201531 11.201531 0.428520 -0.466092 0.838024 25.157895 262.309 ... 0 0 0 0 0 0 0 0 0 0
392 0.0 0.0 CHEMBL12324 11.257704 11.257704 0.462395 -0.462395 0.797990 26.150000 277.344 ... 0 0 0 0 0 0 0 0 0 0
393 0.0 0.0 CHEMBL274107 11.359778 11.359778 0.372211 -0.473241 0.838024 25.157895 262.309 ... 0 0 0 0 0 0 0 0 0 0

5 rows × 213 columns

# Concatenate df_mp4_lb & df_null_lb
df_full = pd.concat([df_null_lb, df_mp4_lb])
df_full
max_phase_lb max_phase molecule_chembl_id MaxAbsEStateIndex MaxEStateIndex MinAbsEStateIndex MinEStateIndex qed SPS MolWt ... fr_sulfide fr_sulfonamd fr_sulfone fr_term_acetylene fr_tetrazole fr_thiazole fr_thiocyan fr_thiophene fr_unbrch_alkane fr_urea
4 0.0 0.0 CHEMBL28 12.020910 12.020910 0.018823 -0.410347 0.631833 10.800000 270.240 ... 0 0 0 0 0 0 0 0 0 0
8 0.0 0.0 CHEMBL8320 10.282778 10.282778 0.120741 -0.120741 0.416681 17.500000 108.096 ... 0 0 0 0 0 0 0 0 0 0
14 0.0 0.0 CHEMBL11833 11.201531 11.201531 0.428520 -0.466092 0.838024 25.157895 262.309 ... 0 0 0 0 0 0 0 0 0 0
392 0.0 0.0 CHEMBL12324 11.257704 11.257704 0.462395 -0.462395 0.797990 26.150000 277.344 ... 0 0 0 0 0 0 0 0 0 0
393 0.0 0.0 CHEMBL274107 11.359778 11.359778 0.372211 -0.473241 0.838024 25.157895 262.309 ... 0 0 0 0 0 0 0 0 0 0
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
1013 1.0 4.0 CHEMBL1025 12.703056 12.703056 0.426312 -4.304784 0.629869 13.000000 184.147 ... 0 0 0 0 0 0 0 0 0 0
1345 1.0 4.0 CHEMBL1128 9.261910 9.261910 0.000000 0.000000 0.608112 10.692308 201.697 ... 0 0 0 0 0 0 0 0 0 0
2028 1.0 4.0 CHEMBL360055 6.476818 6.476818 0.656759 0.656759 0.205822 12.583333 510.828 ... 0 0 0 0 0 0 0 0 0 0
2725 1.0 4.0 CHEMBL1677 6.199769 6.199769 0.000000 0.000000 0.760853 14.250000 234.730 ... 0 0 0 0 0 0 0 0 0 0
3271 1.0 4.0 CHEMBL1200970 2.520809 2.520809 0.000000 0.000000 0.709785 14.000000 348.943 ... 0 0 0 0 0 0 0 0 0 0

5266 rows × 213 columns

  1. Defined X and y variables and trained RF classifier model

Eventually df_full contained 10 active compounds and 5256 inactive compounds (from the value counts).

df_full.value_counts("max_phase_lb")
max_phase_lb
0.0    5256
1.0      10
Name: count, dtype: int64
# Defining X (features) & y (target)
X = df_full.iloc[:, 3:]
y = df_full.iloc[:, 0]
# Checking right data were selected e.g. y as target
y
4       0.0
8       0.0
14      0.0
392     0.0
393     0.0
       ... 
1013    1.0
1345    1.0
2028    1.0
2725    1.0
3271    1.0
Name: max_phase_lb, Length: 5266, dtype: float64
# Convert both X & y to arrays
X = X.to_numpy()
y = y.to_numpy()
# Using train_test_split() this time to split data
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, stratify=y, random_state=1)

After data splitting, a RF classifier was trained with reference to this notebook.

# max_features = "sqrt" by default
rfc = RandomForestClassifier(max_depth=3, random_state=1, max_features="sqrt", oob_score=True)
rfc.fit(X_train, y_train)
RandomForestClassifier(max_depth=3, oob_score=True, random_state=1)
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
  1. Extracted positive prediction probabilities for the testing set and showed confusion matrix with classification metrics
test_probs = rfc.predict_proba(X_test)[:, 1]

Some reference links and explanations for area under the ROC curve and Cohen’s Kappa.

Area under the ROC curve: reference - the area under a curve plot between sensitivity or recall (percent of all 1s classified correctly by a classifier or true positive rate) and specificity (percent of all 0s classified correctly by a classifier, or equivalent to 1 - false positive rate or true negative rate) (Bruce, Bruce, and Gedeck 2020). It is useful for evaluating the performance of a classification model via comparing the true positive rate and false positive rate which are influenced by shifting the decision threshold. Area under the ROC is usually represented as a number ranging from 0 to 1 (1 being a perfect classifier, 0.5 or below meaning a poor, ineffective classifier)

Cohen’s Kappa score: reference - a score that is used to measure the agreement of labelling between two annotators (usually between -1 and 1, the higher the score the better the agreement)

Rather than re-inventing the wheel, the following function code for calculating metrics of the RF model were adapted from this notebook, from GHOST repository. I have only added some comments for clarities, and also added a zero_division parameter for the classification_report to mute the warning message when the results ended up being 0 due to divisions by zero.

def calc_metrics(y_test, test_probs, threshold = 0.5):
    # Target label assigned according to stated decision threshold (default = 0.5)
    # e.g. second annotator (expected label)
    scores = [1 if x>=threshold else 0 for x in test_probs]
    # Calculate area under the ROC curve based on prediction score
    auc = metrics.roc_auc_score(y_test, test_probs)
    # Calculate Cohen's Kappa score
    # e.g. y_test as first annotator (predicted label)
    kappa = metrics.cohen_kappa_score(y_test, scores)
    # Formulate the confusion matrix
    confusion = metrics.confusion_matrix(y_test, scores, labels = list(set(y_test)))
    print('thresh: %.2f, kappa: %.3f, AUC test-set: %.3f'%(threshold, kappa, auc))
    print(confusion)
    print(metrics.classification_report(y_test, scores, zero_division=0.0))
    return

Note: roc_auc_score measures true positive and false positive rates, requiring binary labels (e.g. 0s and 1s) in the data

Then showed confusion matrix along with area under the ROC curve and Cohen’s Kappa.

calc_metrics(y_test, test_probs, threshold = 0.5)
thresh: 0.50, kappa: 0.000, AUC test-set: 0.536
[[1052    0]
 [   2    0]]
              precision    recall  f1-score   support

         0.0       1.00      1.00      1.00      1052
         1.0       0.00      0.00      0.00         2

    accuracy                           1.00      1054
   macro avg       0.50      0.50      0.50      1054
weighted avg       1.00      1.00      1.00      1054

To help with interpreting the confusion matrix, scikit-learn actually had a handy plotting code to visualise the matrix.

ConfusionMatrixDisplay.from_estimator(rfc, X_test, y_test)
plt.show()

It was very obvious that not all of the compounds were classified in the testing set. There were only 1052 compounds classified as true negative, and none in the testing set were labelled as true positive. The likely reason was due to the very imbalanced ratio of actives (only 10 max phase 4 which were labelled as “1” compounds) and inactives (5256 max phase 0 compounds). Besides the imbalanced dataset, the decision threshold was also normally set at 0.5, meaning the classifier was likely going to lose the chance to classify the true positive compounds due to the very skewed ratio of actives to inactives.

  1. Two approaches were used in the GHOST (generalized threshold shifting) paper:

    • Approach 1 (out-of-bag method, more computer efficient, aimed for RF classifiers) based on RDKit blog post or its viewable notebook version via nbviewer

    • Approach 2 led to GHOST procedure with a goal to optimise and shift the decision threshold in any classification methods to catch the minor portion of actives (rather than the major portion of inactives)

    • note: both approaches were shown to be performing similarly in the paper

I only used approach 2 here since the RDKit blog post had already explained approach 1 in depth.

The next step involved extracting prediction probabilities from the RF classifier trained model.

# Get the positive prediction probabilities of the training set
train_probs = rfc.predict_proba(X_train)[:, 1]
  1. Used GHOST strategy in a postprocessing way (note: last post used data re-sampling method in a preprocessing way)

The decision threshold were optimised by using ghostml code via testing various different thresholds, e.g. in spaces of 0.05 that ranged from 0.05 to 0.5. The most optimal threshold would have the most maximised Cohen’s kappa.

# Setting up different decision thresholds
thresholds = np.round(np.arange(0.05,0.55,0.05), 2)
thresholds
array([0.05, 0.1 , 0.15, 0.2 , 0.25, 0.3 , 0.35, 0.4 , 0.45, 0.5 ])
# Looking for the best threshold with the most optimal Cohen's Kappa
new_threshold = ghostml.optimize_threshold_from_predictions(y_train, train_probs, thresholds, ThOpt_metrics = 'ROC')

Using the calc_metrics function again on the newly-found or shifted decision threshold.

calc_metrics(y_train, train_probs, threshold = new_threshold)
thresh: 0.10, kappa: 0.933, AUC test-set: 1.000
[[4204    0]
 [   1    7]]
              precision    recall  f1-score   support

         0.0       1.00      1.00      1.00      4204
         1.0       1.00      0.88      0.93         8

    accuracy                           1.00      4212
   macro avg       1.00      0.94      0.97      4212
weighted avg       1.00      1.00      1.00      4212

Here, after shifting the decision threshold with the most optimal Cohen’s Kappa score, we could see an improved number of compounds labelled within the true negative class (increasing from 1052 to 4204), and more importantly, we could see the true positive class improved from 0 to 7 as well.


Plotting ROC curves

Time for some plots - I’ve shown two different ways to plot ROC curves below.

  1. Using scikit-learn

Testing set ROC curve - obviously, this was not a good classifier with a poor AUC.

RocCurveDisplay.from_predictions(y_test, test_probs, plot_chance_level = True)
<sklearn.metrics._plot.roc_curve.RocCurveDisplay at 0x12ce82710>


Training set ROC curve - this probably looked too good to be true or a textbook-standard ROC curve with AUC at 1.0.

RocCurveDisplay.from_predictions(y_train, train_probs, plot_chance_level = True)
<sklearn.metrics._plot.roc_curve.RocCurveDisplay at 0x12cece890>

  1. Alternative method using matplotlib which reproduced a similar training set ROC plot:
# Looking for true positive rate (tpr), false positive rate (fpr) & threshold
fpr, tpr, thresh = metrics.roc_curve(y_train, train_probs)
# Plotting
plt.figure()
# lw = linewidth
plt.plot(fpr, tpr, lw = 2)
# show random guessing line (threshold = 0.5)
plt.plot([0, 1], [0, 1], color = "g", lw = 2, linestyle="--")
plt.ylim([-0.05, 1.05])
plt.xlim([-0.05, 1.0])
plt.xlabel("specificity")
plt.ylabel("recall")
plt.show()


Downsides and thoughts

I wanted to mention that the testing set used here was most likely not the best ones to be used. There could be many overlaps or similarities between the training and testing sets, since they all came from ChEMBL database. For demonstration and learning purposes, I ended up using similar dataset as last time. Hopefully, I can try other open-source or public drug discovery datasets in the near future.

The other thing to mention was that I should try different molecular fingerprints or descriptors as well, rather than only using RDKit2D, which might lead to different results. I should also probably slowly move onto using multiple datasets or targets in a project, which would likely make things more interesting. On the other hand, I also wanted to avoid this in order to make the topic of interest as clear and simple as possible for me or anyone who’s trying to learn.


Acknowledgements

I’d like to thank Riniker lab again for the GHOST paper, along with all the authors, contributors or developers for all of the software packages mentioned in this post, and also, huge thanks should also go to the authors of the reference notebooks mentioned in the post as well.

Back to top

References

Bruce, Peter, Andrew Bruce, and Peter Gedeck. 2020. Practical Statistics for Data Scientists. https://www.oreilly.com/library/view/practical-statistics-for/9781492072935/.
Esposito, Carmen, Gregory A. Landrum, Nadine Schneider, Nikolaus Stiefl, and Sereina Riniker. 2021. “GHOST: Adjusting the Decision Threshold to Handle Imbalanced Data in Machine Learning.” Journal of Chemical Information and Modeling 61 (6): 2623–40. https://doi.org/10.1021/acs.jcim.1c00160.

Footnotes

  1. h/t: Greg Landrum for his comment on Mastodon for the last RF post (which led to this follow-up post)↩︎

]]> Machine learning projects Tree models Pandas Scikit-learn RDKit ChEMBL database Python https://jhylin.github.io/Data_in_life_blog/posts/17_ML2-2_Random_forest/2_random_forest_classifier.html Tue, 16 Jan 2024 11:00:00 GMT Random forest Jennifer HY Lin https://jhylin.github.io/Data_in_life_blog/posts/17_ML2-2_Random_forest/1_random_forest.html Post updated on 3rd May 2024 - Added comment regarding ImbalancedLearningRegression package (installation tip) & Jupyter notebook link of this post

Quick overview of this post
  • Short introduction of random forest
  • Random forest methods or classes in scikit-learn
  • Random forest regressor model in scikit-learn
  • Training and testing data splits
    • ChEMBL-assigned max phase splits
    • Imbalanced learning regression and max phase splits
  • Scoring metrics of trained models
  • Feature importances in dataset
    • feature_importances_attribute in scikit-learn
    • permutation_importance function in scikit-learn
    • SHAP approach
  • Hyperparameter tuning on number of trees


What is a random forest?

The decision tree model built last time was purely based on one model on its own, which often might not be as accurate or reflective in real-life. To improve the model, the average outcome from multiple models (Breiman 1998) should be considered to see if this would provide a more realistic image. This model averaging approach was also constantly used in our daily lives, for example, using majority votes during decision-making steps.

The same model averaging concept was also used in random forest (Breiman 2001), which as the name suggested, was composed of many decision trees (models) forming a forest. Each tree model would be making its own model prediction. By accruing multiple predictions since we have multiple trees, the average obtained from these predictions would produce one single result in the end. The advantage of this was that it improved the accuracy of the prediction by reducing variances, and also minimised the problem of overfitting the model if it was purely based on one model only (more details in section 1.11.2.1. Random Forests from scikit-learn).

The “random” part of the random forest was introduced in two ways. The first one was via using bootstrap samples, which was also known as bagging or bootstrap aggregating (Bruce, Bruce, and Gedeck 2020), where samples were drawn with replacements within the training datasets for each tree built in the ensemble (also known as the perturb-and-combine technique (Breiman 1998)). While bootstrap sampling was happening, randomness was also incorporated into the training sets at the same time. The second way randomness was introduced was by using a random subset of features for splitting at the nodes, or a full set of features could also be used (although this was generally not recommended). The main goal here was to achieve best splits at each node.


Random forest in scikit-learn

Scikit-learn had two main types of random forest classes - ensemble.RandomForestClassifier() and ensemble.RandomForestRegressor(). When to use which class would depend on the target values. The easiest thing to do was to decide whether the target variables had class labels (binary types or non-continuous variables e.g. yes or no, or other different categories to be assigned) or continuous (numerical) variables, which in this case, if I were to continue using the same dataset from the decision tree series, it would be a continuous variable or feature, pKi, the inhibition constant.

There were also two other alternative random forest methods in scikit-learn, which were ensemble.RandomTreesEmbedding() and ensemble.ExtraTreesClassifier() or ensemble.ExtraTreesRegressor(). The difference for RandomTreesEmbedding() was that it was an unsupervised method that used data transformations (more details from section 1.11.2.6. on “Totally Random Trees Embedding” in scikit-learn). On the other side, there was also an option to use ExtraTreesClassifier() or ExtraTreesRegressor() to generate extremely randomised trees that would go for another level up in randomness (more deatils in section 1.11.2.2. on Extremely Randomized Trees from scikit-learn). The main difference for this type of random forest was that while there was already a random subset of feature selection used (with an intention to select the most discerning features), more randomness were added on top of this by using purely randomly generated splitting rules for picking features at the nodes. The advantage of this type of method was that it would reduce variance and increase the accuracy of the model, but the downside was there might be an increase in bias within the model.


Building a random forest regressor model using scikit-learn

As usual, all the required libraries were imported first.

import pandas as pd
import sklearn
from sklearn.ensemble import RandomForestRegressor

# For imbalanced datasets in regression 
# May need to set env variable (SKLEARN_ALLOW_DEPRECATED_SKLEARN_PACKAGE_INSTALL=True) when installing
# due to package dependency on older sklearn version
import ImbalancedLearningRegression as iblr

# Plots
import matplotlib.pyplot as plt
import seaborn as sns

# Metrics
from sklearn.metrics import mean_squared_error
from sklearn.metrics import r2_score

# Feature importances
# Permutation_importance
from sklearn.inspection import permutation_importance
# SHAP values
import shap

# Hyperparameter tuning
from sklearn.model_selection import cross_val_score, RepeatedKFold

from numpy import mean, std
from natsort import index_natsorted
import numpy as np

# Showing version of scikit-learn used
print(sklearn.__version__)
1.3.2

Importing dataset that was preprocessed from last time - link to data source: first decision tree post.

data = pd.read_csv("ache_2d_chembl.csv")
data.drop(columns = ["Unnamed: 0"], inplace=True)
# Preparing data for compounds with max phase with "NaN" by re-labelling to "null"
data["max_phase"].fillna("null", inplace=True)
data.head()
Setting an item of incompatible dtype is deprecated and will raise in a future error of pandas. Value 'null' has dtype incompatible with float64, please explicitly cast to a compatible dtype first.
molecule_chembl_id pKi max_phase mw fsp3 n_lipinski_hba n_lipinski_hbd n_rings n_hetero_atoms n_heavy_atoms ... sas n_aliphatic_carbocycles n_aliphatic_heterocyles n_aliphatic_rings n_aromatic_carbocycles n_aromatic_heterocyles n_aromatic_rings n_saturated_carbocycles n_saturated_heterocyles n_saturated_rings
0 CHEMBL60745 8.787812 null 245.041526 0.400000 2 1 1 3 13 ... 3.185866 0 0 0 1 0 1 0 0 0
1 CHEMBL208599 10.585027 null 298.123676 0.388889 2 2 4 3 21 ... 4.331775 2 0 2 1 1 2 0 0 0
2 CHEMBL95 6.821023 4.0 198.115698 0.307692 2 2 3 2 15 ... 2.014719 1 0 1 1 1 2 0 0 0
3 CHEMBL173309 7.913640 null 694.539707 0.666667 8 0 2 8 50 ... 2.803680 0 0 0 2 0 2 0 0 0
4 CHEMBL1128 6.698970 4.0 201.092042 0.400000 2 1 1 3 13 ... 3.185866 0 0 0 1 0 1 0 0 0

5 rows × 25 columns


Training/testing splits

Two approaches were used, where one was based purely on max phase split (between max phases null and 4), which was used last time in the decision tree series, and the other one was using the same max phase split but with an ImbalancedLearningRegression method added on top of it.


Preparing training data using max phase split

X variable was set up first from the dataframe, and then converted into a NumPy array, which consisted of the number of samples and number of features. This was kept the same as how it was in the decision tree posts.

Note

It’s usually recommended to copy the original data or dataframe before doing any data manipulations to avoid unnecessary changes to the original dataset (this was not used in the decision tree posts, but since I’m going to use the same set of data again I’m doing it here.)

# X variables (molecular features)
# Make a copy of the original dataframe first
data_mp4 = data.copy()
# Selecting all max phase 4 compounds
data_mp4 = data_mp4[data_mp4["max_phase"] == 4]
print(data_mp4.shape)
data_mp4.head()
(10, 25)
molecule_chembl_id pKi max_phase mw fsp3 n_lipinski_hba n_lipinski_hbd n_rings n_hetero_atoms n_heavy_atoms ... sas n_aliphatic_carbocycles n_aliphatic_heterocyles n_aliphatic_rings n_aromatic_carbocycles n_aromatic_heterocyles n_aromatic_rings n_saturated_carbocycles n_saturated_heterocyles n_saturated_rings
2 CHEMBL95 6.821023 4.0 198.115698 0.307692 2 2 3 2 15 ... 2.014719 1 0 1 1 1 2 0 0 0
4 CHEMBL1128 6.698970 4.0 201.092042 0.400000 2 1 1 3 13 ... 3.185866 0 0 0 1 0 1 0 0 0
6 CHEMBL640 6.000000 4.0 235.168462 0.461538 4 3 1 4 17 ... 1.791687 0 0 0 1 0 1 0 0 0
9 CHEMBL502 7.688246 4.0 379.214744 0.458333 4 0 4 4 28 ... 2.677222 1 1 2 2 0 2 0 1 1
131 CHEMBL481 7.296709 4.0 586.279135 0.515152 10 1 7 10 43 ... 3.632560 0 4 4 1 2 3 0 2 2

5 rows × 25 columns

# Select molecular features for X array (n_samples, n_features)
X_mp4_df = data_mp4[['mw', 'fsp3', 'n_lipinski_hba', 'n_lipinski_hbd', 'n_rings', 'n_hetero_atoms', 'n_heavy_atoms', 'n_rotatable_bonds', 'n_radical_electrons', 'tpsa', 'qed', 'clogp', 'sas', 'n_aliphatic_carbocycles', 'n_aliphatic_heterocyles', 'n_aliphatic_rings', 'n_aromatic_carbocycles', 'n_aromatic_heterocyles', 'n_aromatic_rings', 'n_saturated_carbocycles', 'n_saturated_heterocyles', 'n_saturated_rings']]

print(X_mp4_df.shape)
X_mp4_df.head()
(10, 22)
mw fsp3 n_lipinski_hba n_lipinski_hbd n_rings n_hetero_atoms n_heavy_atoms n_rotatable_bonds n_radical_electrons tpsa ... sas n_aliphatic_carbocycles n_aliphatic_heterocyles n_aliphatic_rings n_aromatic_carbocycles n_aromatic_heterocyles n_aromatic_rings n_saturated_carbocycles n_saturated_heterocyles n_saturated_rings
2 198.115698 0.307692 2 2 3 2 15 0 0 38.91 ... 2.014719 1 0 1 1 1 2 0 0 0
4 201.092042 0.400000 2 1 1 3 13 2 0 20.23 ... 3.185866 0 0 0 1 0 1 0 0 0
6 235.168462 0.461538 4 3 1 4 17 6 0 58.36 ... 1.791687 0 0 0 1 0 1 0 0 0
9 379.214744 0.458333 4 0 4 4 28 6 0 38.77 ... 2.677222 1 1 2 2 0 2 0 1 1
131 586.279135 0.515152 10 1 7 10 43 4 0 114.20 ... 3.632560 0 4 4 1 2 3 0 2 2

5 rows × 22 columns

# Convert X_mp4_df to numpy array
X_mp4 = X_mp4_df.to_numpy()
X_mp4
array([[ 1.98115698e+02,  3.07692308e-01,  2.00000000e+00,
         2.00000000e+00,  3.00000000e+00,  2.00000000e+00,
         1.50000000e+01,  0.00000000e+00,  0.00000000e+00,
         3.89100000e+01,  7.06488238e-01,  2.69580000e+00,
         2.01471913e+00,  1.00000000e+00,  0.00000000e+00,
         1.00000000e+00,  1.00000000e+00,  1.00000000e+00,
         2.00000000e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00],
       [ 2.01092042e+02,  4.00000000e-01,  2.00000000e+00,
         1.00000000e+00,  1.00000000e+00,  3.00000000e+00,
         1.30000000e+01,  2.00000000e+00,  0.00000000e+00,
         2.02300000e+01,  6.08112327e-01, -1.01700000e+00,
         3.18586632e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00,  1.00000000e+00,  0.00000000e+00,
         1.00000000e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00],
       [ 2.35168462e+02,  4.61538462e-01,  4.00000000e+00,
         3.00000000e+00,  1.00000000e+00,  4.00000000e+00,
         1.70000000e+01,  6.00000000e+00,  0.00000000e+00,
         5.83600000e+01,  7.31539693e-01,  1.34040000e+00,
         1.79168720e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00,  1.00000000e+00,  0.00000000e+00,
         1.00000000e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00],
       [ 3.79214744e+02,  4.58333333e-01,  4.00000000e+00,
         0.00000000e+00,  4.00000000e+00,  4.00000000e+00,
         2.80000000e+01,  6.00000000e+00,  0.00000000e+00,
         3.87700000e+01,  7.47461492e-01,  4.36110000e+00,
         2.67722173e+00,  1.00000000e+00,  1.00000000e+00,
         2.00000000e+00,  2.00000000e+00,  0.00000000e+00,
         2.00000000e+00,  0.00000000e+00,  1.00000000e+00,
         1.00000000e+00],
       [ 5.86279135e+02,  5.15151515e-01,  1.00000000e+01,
         1.00000000e+00,  7.00000000e+00,  1.00000000e+01,
         4.30000000e+01,  4.00000000e+00,  0.00000000e+00,
         1.14200000e+02,  3.55955569e-01,  4.09110000e+00,
         3.63256044e+00,  0.00000000e+00,  4.00000000e+00,
         4.00000000e+00,  1.00000000e+00,  2.00000000e+00,
         3.00000000e+00,  0.00000000e+00,  2.00000000e+00,
         2.00000000e+00],
       [ 5.10461822e+02,  8.00000000e-01,  6.00000000e+00,
         0.00000000e+00,  1.00000000e+00,  6.00000000e+00,
         3.60000000e+01,  2.10000000e+01,  0.00000000e+00,
         2.76900000e+01,  2.05822189e-01,  5.45250000e+00,
         3.25765349e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00,  1.00000000e+00,  0.00000000e+00,
         1.00000000e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00],
       [ 1.84066459e+02,  1.00000000e+00,  3.00000000e+00,
         0.00000000e+00,  0.00000000e+00,  5.00000000e+00,
         1.10000000e+01,  4.00000000e+00,  0.00000000e+00,
         3.55300000e+01,  6.29869319e-01,  2.91400000e+00,
         3.34514393e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00],
       [ 2.87152144e+02,  5.29411765e-01,  4.00000000e+00,
         1.00000000e+00,  4.00000000e+00,  4.00000000e+00,
         2.10000000e+01,  1.00000000e+00,  0.00000000e+00,
         4.19300000e+01,  8.00524269e-01,  1.85030000e+00,
         4.22684283e+00,  1.00000000e+00,  2.00000000e+00,
         3.00000000e+00,  1.00000000e+00,  0.00000000e+00,
         1.00000000e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00],
       [ 3.48142697e+02,  3.68421053e-01,  2.00000000e+00,
         0.00000000e+00,  3.00000000e+00,  4.00000000e+00,
         2.30000000e+01,  5.00000000e+00,  0.00000000e+00,
         6.48000000e+00,  7.09785317e-01,  5.44140000e+00,
         4.22359068e+00,  0.00000000e+00,  1.00000000e+00,
         1.00000000e+00,  2.00000000e+00,  0.00000000e+00,
         2.00000000e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00],
       [ 2.34092376e+02,  3.07692308e-01,  2.00000000e+00,
         2.00000000e+00,  3.00000000e+00,  3.00000000e+00,
         1.60000000e+01,  0.00000000e+00,  0.00000000e+00,
         3.89100000e+01,  7.60853221e-01,  3.11760000e+00,
         3.21871482e+00,  1.00000000e+00,  0.00000000e+00,
         1.00000000e+00,  1.00000000e+00,  1.00000000e+00,
         2.00000000e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00]])

Again, y variable was arranged via the dataframe as well, and converted into a NumPy array. It consisted of the number of samples only as this was the target variable.

# y array (n_samples) - target outcome pKi
y_mp4_df = data_mp4["pKi"]
y_mp4_df
2      6.821023
4      6.698970
6      6.000000
9      7.688246
131    7.296709
133    4.431798
160    5.221849
171    6.522879
180    4.607303
195    6.995679
Name: pKi, dtype: float64
# Convert y_mp4_df to numpy array
y_mp4 = y_mp4_df.to_numpy()
y_mp4
array([6.82102305, 6.69897   , 6.        , 7.68824614, 7.29670862,
       4.43179828, 5.22184875, 6.52287875, 4.60730305, 6.99567863])


Training model using max phase split only

Both X and y variables were used to fit the RandomForestRegressor() estimator.

# n_estimators = 100 by default
# note: if wanting to use whole dataset - switch off "bootstrap" parameter by using "False"
rfreg = RandomForestRegressor(max_depth=3, random_state=1, max_features=0.3)
rfreg.fit(X_mp4, y_mp4)
RandomForestRegressor(max_depth=3, max_features=0.3, random_state=1)
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.


Preparing testing data using max phase split only

Testing data was mainly based on compounds with max phase assigned as “0” or “null” after I renamed it above.

data_mp_null = data.copy()
# Selecting all max phase "null" compounds
data_mp_null = data_mp_null[data_mp_null["max_phase"] == "null"]
print(data_mp_null.shape)
data_mp_null.head() 
(466, 25)
molecule_chembl_id pKi max_phase mw fsp3 n_lipinski_hba n_lipinski_hbd n_rings n_hetero_atoms n_heavy_atoms ... sas n_aliphatic_carbocycles n_aliphatic_heterocyles n_aliphatic_rings n_aromatic_carbocycles n_aromatic_heterocyles n_aromatic_rings n_saturated_carbocycles n_saturated_heterocyles n_saturated_rings
0 CHEMBL60745 8.787812 null 245.041526 0.400000 2 1 1 3 13 ... 3.185866 0 0 0 1 0 1 0 0 0
1 CHEMBL208599 10.585027 null 298.123676 0.388889 2 2 4 3 21 ... 4.331775 2 0 2 1 1 2 0 0 0
3 CHEMBL173309 7.913640 null 694.539707 0.666667 8 0 2 8 50 ... 2.803680 0 0 0 2 0 2 0 0 0
5 CHEMBL102226 4.698970 null 297.152928 0.923077 3 0 0 5 18 ... 2.965170 0 0 0 0 0 0 0 0 0
7 CHEMBL103873 5.698970 null 269.121628 0.909091 3 0 0 5 16 ... 3.097106 0 0 0 0 0 0 0 0 0

5 rows × 25 columns

# Set up X test variable with the same molecular features
X_mp_test_df = data_mp_null[['mw', 'fsp3', 'n_lipinski_hba', 'n_lipinski_hbd', 'n_rings', 'n_hetero_atoms', 'n_heavy_atoms', 'n_rotatable_bonds', 'n_radical_electrons', 'tpsa', 'qed', 'clogp', 'sas', 'n_aliphatic_carbocycles', 'n_aliphatic_heterocyles', 'n_aliphatic_rings', 'n_aromatic_carbocycles', 'n_aromatic_heterocyles', 'n_aromatic_rings', 'n_saturated_carbocycles', 'n_saturated_heterocyles', 'n_saturated_rings']]

# Convert X test variables from df to arrays
X_mp_test = X_mp_test_df.to_numpy()

X_mp_test
array([[2.45041526e+02, 4.00000000e-01, 2.00000000e+00, ...,
        0.00000000e+00, 0.00000000e+00, 0.00000000e+00],
       [2.98123676e+02, 3.88888889e-01, 2.00000000e+00, ...,
        0.00000000e+00, 0.00000000e+00, 0.00000000e+00],
       [6.94539707e+02, 6.66666667e-01, 8.00000000e+00, ...,
        0.00000000e+00, 0.00000000e+00, 0.00000000e+00],
       ...,
       [3.11152144e+02, 3.15789474e-01, 4.00000000e+00, ...,
        0.00000000e+00, 0.00000000e+00, 0.00000000e+00],
       [3.68096076e+02, 9.23076923e-01, 4.00000000e+00, ...,
        0.00000000e+00, 2.00000000e+00, 2.00000000e+00],
       [2.46136828e+02, 5.00000000e-01, 4.00000000e+00, ...,
        0.00000000e+00, 3.00000000e+00, 3.00000000e+00]])


Training/testing splits using ImbalancedLearningRegression and max phase splits

I didn’t really pay a lot of attentions when I was doing data splits in the decision tree series, as my main focus was on building a single tree in order to fully understand and see what could be derived from just one tree. Now, when I reached this series on random forest, I realised I forgot to mention in the last series that data splitting was actually very crucial on model performance and could influence outcome predictions. It could also become quite complicated as more approaches were available to split the data. Also, the way the data was splitted could produce different outcomes.

After I’ve splitted the same dataset based on compounds’ max phase assignments in ChEMBL and also fitted the training data on the random forest regressor, I went back and noticed that the training and testing data were very imbalanced and I probably should do something about it before fitting them onto another model.

At this stage, I went further to look into whether imbalanced datasets should be addressed in regression tasks, and did a surface search online. So based on common ML concensus, addressing imbalanced datasets were more applicable to classification tasks (e.g. binary labels or multi-class labels), rather than regression problems. However, recent ML research looked into the issue of imbalanced datasets in regression. This blog post mentioned a few studies that looked into this type of problem, and I thought they were very interesting and worth a mention at least. One of them that I’ve looked into was SMOTER, which was based on synthetic minority over-sampling technique (SMOTE)(Chawla et al. 2002), and was named this way because it was basically a SMOTE for regression (hence SMOTER)(Torgo et al. 2013). Synthetic minority over-sampling technique for regression with Gaussian noise (SMOGN)(Kunz 2020) was another technique that was built upon SMOTER, but with Gaussian noises added. This has subsequently led me to ImbalancedLearningRegression library (Wu, Kunz, and Branco 2022), which was a variation of SMOGN. This was the one used on my imbalanced dataset, shown in the section below.

A simple flow diagram was drawn below showing the evolution of different techniques when dealing with imbalanced datasets in classification (SMOTE) and regression (SMOTER, SMOGN and ImbalancedLearningRegression):

flowchart LR
  A(SMOTE) --> B(SMOTER)
  B --> C(SMOGN)
  C --> D(ImbalancedLearningRegression)

GitHub repository for ImbalancedLearningRegression package is available here, with its documentation available here.

Also, I just wanted to mention that these were not the only techniques available for treating imbalanced datasets in regression, as there were other ones in the literature and most likely more are being developed currently, but I only had time to cover these here for now.

I also would like to mention another really useful open-source resource for treating imbalanced datasets in classifications since I did not use it in this post due to the problem being more of a regression one than a classification one - imbalance-learn library.

# Original dataset - checking shape again
print(data.shape)
data.head()
(481, 25)
molecule_chembl_id pKi max_phase mw fsp3 n_lipinski_hba n_lipinski_hbd n_rings n_hetero_atoms n_heavy_atoms ... sas n_aliphatic_carbocycles n_aliphatic_heterocyles n_aliphatic_rings n_aromatic_carbocycles n_aromatic_heterocyles n_aromatic_rings n_saturated_carbocycles n_saturated_heterocyles n_saturated_rings
0 CHEMBL60745 8.787812 null 245.041526 0.400000 2 1 1 3 13 ... 3.185866 0 0 0 1 0 1 0 0 0
1 CHEMBL208599 10.585027 null 298.123676 0.388889 2 2 4 3 21 ... 4.331775 2 0 2 1 1 2 0 0 0
2 CHEMBL95 6.821023 4.0 198.115698 0.307692 2 2 3 2 15 ... 2.014719 1 0 1 1 1 2 0 0 0
3 CHEMBL173309 7.913640 null 694.539707 0.666667 8 0 2 8 50 ... 2.803680 0 0 0 2 0 2 0 0 0
4 CHEMBL1128 6.698970 4.0 201.092042 0.400000 2 1 1 3 13 ... 3.185866 0 0 0 1 0 1 0 0 0

5 rows × 25 columns

So my little test on using ImbalancedLearningRegression package started from below.

iblr_data = data.copy()

# Introducing Gaussian noise for data sampling
data_gn = iblr.gn(data = iblr_data, y = "pKi", pert = 1)
print(data_gn.shape)
synth_matrix:   0%|          | 0/58 [00:00<?, ?it/s]
synth_matrix:  14%|#3        | 8/58 [00:00<00:00, 75.97it/s]
synth_matrix:  31%|###1      | 18/58 [00:00<00:00, 84.29it/s]
synth_matrix:  47%|####6     | 27/58 [00:00<00:00, 85.90it/s]
synth_matrix:  62%|######2   | 36/58 [00:00<00:00, 83.02it/s]
synth_matrix:  78%|#######7  | 45/58 [00:00<00:00, 83.24it/s]
synth_matrix:  93%|#########3| 54/58 [00:00<00:00, 84.17it/s]
synth_matrix: 100%|##########| 58/58 [00:00<00:00, 84.10it/s]
r_index:   0%|          | 0/8 [00:00<?, ?it/s]
r_index: 100%|##########| 8/8 [00:00<00:00, 267.33it/s]
(480, 25)
# Followed by max phase split, where max phase 4 = training dataset
data_gn_mp4 = data_gn[data_gn["max_phase"] == 4]
data_gn_mp4
print(data_gn_mp4.shape)
(7, 25)
# Also splitted max phase null compounds = testing dataset
data_gn_mp_null = data_gn[data_gn["max_phase"] == "null"]
data_gn_mp_null
print(data_gn_mp_null.shape)
(465, 25)

There were several different sampling techniques in ImbalancedLearningRegression package. I’ve only tried random over-sampling, under-sampling and Gaussian noise, but there were also other ones such as SMOTE and ADASYN (in over-sampling technique) or condensed nearest neighbor, Tomeklinks and edited nearest neightbour (in under-sampling technique) that I haven’t used.

Random over-sampling actually oversampled the max phase null compounds (sample size increased), while keeping all 10 max phase 4 compounds. Under-sampling removed all of the max phase 4 compounds (which was most likely not the best option, since I was aiming to use them as training compounds), with max phase null compounds also reduced in size too. Due to post length, I did not show the code for random over-sampling and under-sampling, but for people who are interested, I think it would be interesting to test them out.

I ended up using Gauissian noise sampling and it reduced max phase 4 compounds slightly, and increased the max phase null compounds a little bit too, which seemed to be the most balanced data sampling at the first try. (Note: as stated from the documentation for ImbalancedLearningRegression package, missing values within features would be removed automatically, I’ve taken care of this in my last series of posts so no difference were observed here.)

The change in the distribution of pKi values for the Gaussian noise sampling method between the original and sample-modified datasets could be seen in the kernel density estimate plot below. The modified dataset had a flatter target density curve than the original density plot, which was more concentrated and peaked between pKi values of 6 and 8. The range of pKi values for the ten max phase 4 compounds collected was between 4 and 8.

Plot reference

# Quick look at how the pKi values differed 
# after applying Gaussian noise sampling to dataset
# Plot target variable, pKi distributions
sns.kdeplot(data["pKi"], label = "Original")
sns.kdeplot(data_gn["pKi"], label = "Modified")
plt.legend(labels = ["Original", "Modified"])
<matplotlib.legend.Legend at 0x13b87ae60>

Next, the modified ImbalancedLearningRegression-Gaussian noise (iblr-gn) training data was converted into a NumPy array.

# Select molecular features for X variable
X_mp4_gn_df = data_gn_mp4[['mw', 'fsp3', 'n_lipinski_hba', 'n_lipinski_hbd', 'n_rings', 'n_hetero_atoms', 'n_heavy_atoms', 'n_rotatable_bonds', 'n_radical_electrons', 'tpsa', 'qed', 'clogp', 'sas', 'n_aliphatic_carbocycles', 'n_aliphatic_heterocyles', 'n_aliphatic_rings', 'n_aromatic_carbocycles', 'n_aromatic_heterocyles', 'n_aromatic_rings', 'n_saturated_carbocycles', 'n_saturated_heterocyles', 'n_saturated_rings']]

print(X_mp4_gn_df.shape)
X_mp4_gn_df.head()
(7, 22)
mw fsp3 n_lipinski_hba n_lipinski_hbd n_rings n_hetero_atoms n_heavy_atoms n_rotatable_bonds n_radical_electrons tpsa ... sas n_aliphatic_carbocycles n_aliphatic_heterocyles n_aliphatic_rings n_aromatic_carbocycles n_aromatic_heterocyles n_aromatic_rings n_saturated_carbocycles n_saturated_heterocyles n_saturated_rings
244 348.142697 0.368421 2.0 0.0 3.0 4.0 23.0 5.0 0 6.48 ... 4.223591 0.0 1.0 1.0 2.0 0.0 2.0 0.0 0.0 0.0
247 201.092042 0.400000 2.0 1.0 1.0 3.0 13.0 2.0 0 20.23 ... 3.185866 0.0 0.0 0.0 1.0 0.0 1.0 0.0 0.0 0.0
266 184.066459 1.000000 3.0 0.0 0.0 5.0 11.0 4.0 0 35.53 ... 3.345144 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
365 198.115698 0.307692 2.0 2.0 3.0 2.0 15.0 0.0 0 38.91 ... 2.014719 1.0 0.0 1.0 1.0 1.0 2.0 0.0 0.0 0.0
385 235.168462 0.461538 4.0 3.0 1.0 4.0 17.0 6.0 0 58.36 ... 1.791687 0.0 0.0 0.0 1.0 0.0 1.0 0.0 0.0 0.0

5 rows × 22 columns

X_mp4_gn = X_mp4_gn_df.to_numpy()

Similarly, this was also applied to the target y variable in the iblr-gn dataset.

# y variable (target outcome - pKi)
y_mp4_gn_df = data_gn_mp4["pKi"]

y_mp4_gn = y_mp4_gn_df.to_numpy()
y_mp4_gn
array([4.60730305, 6.69897   , 5.22184875, 6.82102305, 6.        ,
       7.68824614, 6.99567863])

Then the iblr-gn training data were fitted onto another random forest regressor model.

# n_estimators = 100 by default
# note: if wanting to use whole dataset - switch off "bootstrap" parameter by using "False"
rfreg_gn = RandomForestRegressor(max_depth=3, random_state=1, max_features=0.3)
rfreg_gn.fit(X_mp4_gn, y_mp4_gn)
RandomForestRegressor(max_depth=3, max_features=0.3, random_state=1)
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

Modified iblr-gn testing data were also prepared and converted into a NumPy array.

# Set up X test variable with the same molecular features
X_mp_gn_test_df = data_gn_mp_null[['mw', 'fsp3', 'n_lipinski_hba', 'n_lipinski_hbd', 'n_rings', 'n_hetero_atoms', 'n_heavy_atoms', 'n_rotatable_bonds', 'n_radical_electrons', 'tpsa', 'qed', 'clogp', 'sas', 'n_aliphatic_carbocycles', 'n_aliphatic_heterocyles', 'n_aliphatic_rings', 'n_aromatic_carbocycles', 'n_aromatic_heterocyles', 'n_aromatic_rings', 'n_saturated_carbocycles', 'n_saturated_heterocyles', 'n_saturated_rings']]

# Convert X test variables from df to arrays
X_mp_gn_test = X_mp_gn_test_df.to_numpy()

X_mp_gn_test
array([[5.80224119e+02, 2.64705882e-01, 7.00000000e+00, ...,
        0.00000000e+00, 0.00000000e+00, 0.00000000e+00],
       [3.77210327e+02, 3.91304348e-01, 5.00000000e+00, ...,
        0.00000000e+00, 1.00000000e+00, 1.00000000e+00],
       [5.24297368e+02, 4.54545455e-01, 4.00000000e+00, ...,
        0.00000000e+00, 0.00000000e+00, 0.00000000e+00],
       ...,
       [3.37167794e+02, 2.85714286e-01, 4.00000000e+00, ...,
        0.00000000e+00, 0.00000000e+00, 0.00000000e+00],
       [5.09129751e+02, 4.16666667e-01, 9.00000000e+00, ...,
        0.00000000e+00, 0.00000000e+00, 0.00000000e+00],
       [3.40189926e+02, 4.21052632e-01, 6.00000000e+00, ...,
        0.00000000e+00, 0.00000000e+00, 0.00000000e+00]])


Using trained model for prediction on testing data

Predicting max phase-splitted data only.

# Predict pKi values for the compounds with "null" max phase
# using the training model rfreg 
# Uncomment code below to print prediction result
#print(rfreg.predict(X_mp_test))

# or use:
y_mp_test = rfreg.predict(X_mp_test)

Predicting iblr-gn data with max phase splits.

y_mp_gn_test = rfreg_gn.predict(X_mp_gn_test)


Scoring and metrics of trained models

Checking model accuracy for both training and testing datasets was recommended to take place before moving onto discovering feature importances. A scikit-learn explanation for this could be found in the section on “Permutation feature importance”. So the accuracy scores for the model were shown below.

# Training set accuracy
print(f"Random forest regressor training accuracy: {rfreg.score(X_mp4, y_mp4):.2f}")

# Testing set accuracy
print(f"Random forest regressor testing accuracy: {rfreg.score(X_mp_test, y_mp_test):.2f}")
Random forest regressor training accuracy: 0.86
Random forest regressor testing accuracy: 1.00

It looked like both the training and testing accuracies for the random forest regressor model (rfreg) were quite high, meaning that the model was able to remember the molecular features well from the training set (the tiny sample of 10 compounds), and the model was able to apply them to the testing set (which should contain about 400 or so compounds) as well, in order to make predictions on the target value of pKi. This has somewhat confirmed that the model was indeed making predictions, rather than not making any predictions at all, which meant there might be no point in finding out which features were important in the data. Therefore, we could now move onto processing the feature importances to fill in the bigger story i.e. which features were more pivotal towards influencing pKi values of approved drugs targeting acetylcholinesterase (AChE).

Similar model accuracy scores were also generated for the iblr-gn modified dataset, which appeared to follow a similar pattern as the max phase-splitted dataset.

# iblr-Gaussian noise & max phase splitted data
# Training set accuracy
print(f"Random forest regressor training accuracy: {rfreg_gn.score(X_mp4_gn, y_mp4_gn):.2f}")

# Testing set accuracy
print(f"Random forest regressor testing accuracy: {rfreg_gn.score(X_mp_gn_test, y_mp_gn_test):.2f}")
Random forest regressor training accuracy: 0.79
Random forest regressor testing accuracy: 1.00

Now, setting up the y_true, which was the acutal pKi values of the testing set, and were converted into a NumPy array too.

y_true = data_mp_null["pKi"]
y_true = y_true.to_numpy(copy=True)

I also found out the mean squared error (MSE) between y_true (actual max phase null compounds’ pKi values) and y_pred (predicted max phase null compounds’ pKi values). When MSE was closer to zero, the better the model was, meaning less errors were present.

Some references that might help with explaining MSE:

# For max phase splitted dataset only
mean_squared_error(y_true, y_mp_test)
2.3988097789702505

When R2 (coefficient of determination) was closer to 1, the better the model is, with a usual range between 0 and 1 (Bruce, Bruce, and Gedeck 2020). If it was negative, then the model might not be performing as well as expected. However, there could be exceptions as other model evaluation methods should also be interpreted together with R2 (a poor R2 might not be wholly indicating it’s a poor model).

Some references that might help with understanding R2:

# For max phase splitted dataset only
r2_score(y_true, y_mp_test)
-0.16228227953132635

Because the data was re-sampled in a iblr-gn way, the size of array would be different from the original dataset, so here I’ve specifically grabbed pKi values from the iblr-gn modified data to get the actual pKi values for the max phase null compounds.

y_true_gn = data_gn_mp_null["pKi"]
y_true_gn = y_true_gn.to_numpy(copy=True)
# MSE for iblr-gn data
mean_squared_error(y_true_gn, y_mp_gn_test)
5.7895732090189185
# R squared for iblr-gn data
r2_score(y_true_gn, y_mp_gn_test)
-0.7425920410726885

Well, it appeared iblr-gn dataset might not offer much advantage than the original max phase splitted method. However, even the max phase splitted method wasn’t that great either, but it might still be interesting to find out which features were important in relation to the pKi values.


Feature importances

There were two types of feature importances available in scikit-learn, which I’ve described below. I’ve also added a Shapley additive explanations (SHAP) approach to this section as well to show different visualisation styles for feature importances on the same set of data.


feature_importances_ attribute from scikit-learn

The impurity-based feature importances (also known as Gini importance) were shown below.

# Compute feature importances on rfreg training model
feature_imp = rfreg.feature_importances_
# Check what feature_imp looks like (an array)
feature_imp
array([0.0396524 , 0.06608031, 0.01129602, 0.09707957, 0.03625411,
       0.04393835, 0.05287614, 0.03339396, 0.        , 0.14225505,
       0.03562854, 0.16945464, 0.04450665, 0.02810539, 0.01437198,
       0.0220527 , 0.02136604, 0.00803076, 0.04866529, 0.        ,
       0.044468  , 0.04052409])

I decided to write a function to convert a NumPy array into a plot below as this was also needed in the next section.

# Function to convert array to df leading to plots 
# - for use in feature_importances_ & permutation_importance

def feat_imp_plot(feat_imp_array, X_df):

    """
    Function to convert feature importance array into a dataframe, 
    which is then used to plot a bar graph 
    to show the feature importance ranking in the random forest model for the dataset used.

    feat_imp_array is the array obtained from the feature_importances_ attribute, 
    after having a estimator/model fitted.

    X_df is the dataframe for the X variable, 
    where the feature column names will be used in the plot.
    """

    # Convert the feat_imp array into dataframe
    feat_imp_df = pd.DataFrame(feat_imp_array)
    
    # Obtain feature names via column names of dataframe
    # Rename the index as "features"
    feature = X_df.columns.rename("features")

    # Convert the index to dataframe
    feature_name_df = feature.to_frame(index = False)

    # Concatenate feature_imp_df & feature_name_df
    feature_df = pd.concat(
        [feat_imp_df, feature_name_df], 
        axis=1
        ).rename(
            # Rename the column for feature importances
            columns = {0: "feature_importances"}
            ).sort_values(
                # Sort values of feature importances in descending order
                "feature_importances", ascending=False
                )
    
    # Seaborn bar plot
    sns.barplot(
        feature_df, 
        x = "feature_importances", 
        y = "features")
# Testing feat_imp_plot function
feat_imp_plot(feature_imp, X_mp4_df)

An alternative way to plot was via Matplotlib directly (note: Seaborn was built based on Matplotlib, so the plots were pretty similar). The code below were probably a bit more straightforward but without axes named and the values were not sorted (only as an example but more code could be added to do this).

# Matplotlib plot
from matplotlib import pyplot as plt
plt.barh(X_mp4_df.columns, rfreg.feature_importances_)
<BarContainer object of 22 artists>


permutation_importance function from scikit-learn

There were known issues with the built-in feature_importances_ attribute in scikit-learn. As quoted from scikit-learn on feature importance evaluation:

… The impurity-based feature importances computed on tree-based models suffer from two flaws that can lead to misleading conclusions. First they are computed on statistics derived from the training dataset and therefore do not necessarily inform us on which features are most important to make good predictions on held-out dataset. Secondly, they favor high cardinality features, that is features with many unique values. Permutation feature importance is an alternative to impurity-based feature importance that does not suffer from these flaws. …

So I’ve also tried the permutation_importance function (a model-agnostic method).

perm_result = permutation_importance(rfreg, X_mp_test, y_mp_test, n_repeats=10, random_state=1, n_jobs=2)

# Checking data type of perm_result
type(perm_result)
sklearn.utils._bunch.Bunch

It normally returns a dictionary-like objects (e.g. Bunch) with the following 3 attributes:

  • importances_mean (mean of feature importances)
  • importances_std (standard deviation of feature importances)
  • importances (raw permutation/feature importances scores)

For details on these attributes, this scikit-learn link will add a bit more explanations.

I decided to only use importances_mean for now.

perm_imp = perm_result.importances_mean

# Confirm it produces an array
type(perm_imp)
numpy.ndarray
# Using the function feat_imp_plot() on perm_imp result to show plot
feat_imp_plot(perm_imp, X_mp4_df)

It generated a different feature importances ranking (if looking at top 6 features), although somewhat similar to the previous one.


SHAP approach

SHAP values (Lundberg et al. 2020), (Shapley et al. 1953) were used here to provide another way to figure out feature importances. The GitHub repository for this SHAP approach could be accessed here.

SHAP’s TreeExplainer() was based on Tree SHAP algorithms (Lundberg et al. 2020), and was used to show and explain feature importances within tree models. It could also be extended to boosted tree models such as LightGBM and XGBoost and also other tree models (as explained by the GitHub repository README.md and its documentation link provided). It was also a model-agnostic method, which could be quite handy.

Other reference

shap_explainer = shap.TreeExplainer(rfreg)

# X_test needs to be a dataframe (not numpy array)
# otherwise feature names won't show in plot
shap_values = shap_explainer.shap_values(X_mp_test_df)

# Horizontal bar plot
shap.summary_plot(shap_values, X_mp_test_df, plot_type = "bar")

Dot plot version:

shap.summary_plot(shap_values, X_mp_test_df)

Violin plot:

shap.summary_plot(shap_values, X_mp_test_df, plot_type = "violin")

# Alternative plot option: "layered_violin"


Hyperparameter tuning

An example was shown below on tuning the number of trees (n_estimators) used in the random forest model.

# Function code adapted with thanks from ML Mastery 
# https://machinelearningmastery.com/random-forest-ensemble-in-python/

# ---Evaluate a list of models with different number of trees---

# Define dataset by using the same training dataset as above
X, y = X_mp4, y_mp4

# Define function to generate a list of models with different no. of trees
def models():
    # Create empty dictionary (key, value pairs) for models
    models = dict()
    # Test different number of trees to evaluate
    no_trees = [50, 100, 250, 500, 1000]
    for n in no_trees:
        models[str(n)] = RandomForestRegressor(n_estimators=n)
    return models


# Define function to evaluate a single model using cross-validation
def evaluate(model, X, y):

    # RepeatedStratifiedKFold usually for binary or multi-class labels 
    # - ref link: https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.KFold.html#sklearn.model_selection.KFold
    # so using ReaptedKFold instead
    cross_val = RepeatedKFold(n_splits=10, n_repeats=15, random_state=1)
    # Run evaluation process & collect cv scores
    # Since estimator/model was based on DecisionTreeRegressor, 
    # using neg_mean_squared_error metric
    # n_jobs = -1 meaning using all processors to run jobs in parallel
    scores = cross_val_score(model, X, y, scoring="neg_mean_squared_error", cv=cross_val, n_jobs=-1)
    return scores


# Evaluate results
# Run models with different RepeatedKFold & different no. of tress
# with results shown as diff. trees with calculated mean cv scores & std

# Obtain diff. models with diff. trees via models function
models = models()

# Create empty lists for results & names
results, names = list(), list()

# Create a for loop to iterate through the list of diff. models
for name, model in models.items():
    # Run the cross validation scores via evaluate function
    scores = evaluate(model, X, y)
    # Collect results
    results.append(scores)
    # Collect names (different no. of trees)
    names.append(name)
    # Show the average mean squared errors and corresponding standard deviations 
    # for each model with diff. no. of trees
    print((name, mean(scores), std(scores)))
('50', -1.6470594650953017, 1.6444082604560304)
('100', -1.6995136024743887, 1.6797340671624852)
('250', -1.6716290617106646, 1.6236808789148038)
('500', -1.645981936868625, 1.615445700037851)
('1000', -1.6532678610618743, 1.604259597928101)

The negated version of the mean squared error (neg_mean_squared_error) was due to how the scoring parameter source code was written in scikit-learn. It was written this way to take into account of both scoring and loss functions (links provided below for further explanations). All scoring metrics could be accessed here for scikit-learn.

Reference links to help with understanding neg_mean_squared_error:

  1. scikit-learn source code

  2. StackOverflow answer

Also, the random forest algorithm was stochastic in nature, meaning that every time hyperparameter tuning took place, it would generate different scores due to random bootstrap sampling. The best approach to evaluate model performance during the cross-validation process was to use the average outcome from several runs of cross-validations, then fit the hyperparameters on a final model, or getting several final models ready and then obtaining the average from these models instead.

Below was a version of boxplot plotted using Matplotlib showing the differences in the distributions of the cross validation scores and mean squared errors between different number of trees.

plt.boxplot(results, labels=names, showmeans=True)
plt.show()

To plot this in Seaborn, I had to prepare the data slightly differently to achieve a different version of the boxplot. Matplotlib was a bit more straightforward to use without these steps.

I also used natural sort to sort numerical values (GitHub repository). Otherwise, if using sort_values() only, it would only sort the numbers in lexicographical order (i.e. by first digit only), which was not able to show the tree numbers in ascending order.

# Combine results & names lists into dataframe
cv_results = pd.DataFrame(results, index = [names])
# Reset index and rename the number of trees column
cv_results = cv_results.reset_index().rename(columns={"level_0": "Number_of_trees"})
# Melt the dataframe by number of trees column
cv_results = cv_results.melt(id_vars="Number_of_trees")
# Sort by the number of trees column
cv_results = cv_results.sort_values(
    by="Number_of_trees",
    key=lambda x: np.argsort(index_natsorted(cv_results["Number_of_trees"]))
)
# Seaborn boxplot
sns.boxplot(cv_results, x="Number_of_trees", y="value", showmeans=True)
<Axes: xlabel='Number_of_trees', ylabel='value'>

The Seaborn boxplot shown should be very similar to the Matplotlib one.

Other hyperparameters that could be tuned included:

  • tree depths (max_depth)

  • number of samples (max_samples)

  • number of features (max_features) - I didn’t use RDKit to generate molecular features for this post (Datamol version was used instead) which would provide around 209 at least (trying to keep the post at a readable length), but I think this might be a better option when doing cross-validations in model evaluations

  • number of nodes (max_leaf_nodes)

I’ve decided not to code for these other hyperparameters in the cross-validation step due to length of post (the function code used in cross-validation above could be further adapted to cater for other hyperparameters mentioned here), but they should be looked into if doing full-scale and comprehensive ML using the ensemble random forest algorithm.


Final words

Random forest was known to be a black-box ML algorithm (Bruce, Bruce, and Gedeck 2020), which was completely different from the white-box ML style revealed in decision tree graphs. Feature importances was therefore crucial to shed some lights and remove some layers of the black-box nature in random forest by showing which features were contributing towards model accuracy by ranking features used to train the model. Cross-validation was also vital to avoid over-fitting (which was more applicable to depth of trees), although in some other cases (e.g. number of trees), it was mentioned that it was unlikely the model would be overfitted. Other options available in scikit-learn ensemble methods that I didn’t get time to try were using voting classifier/regressor and stacking models to reduce biases in models, which might be very useful in other cases.

Few things I’ve thought of that I could try to improve what I did here was that I should really look for a different set of testing data, rather than using the max phase splits, which was not that ideal. However, as a lot of us are aware, good drug discovery data are hard to come by (a long-standing and complicated problem), I probably need some luck while looking for a different set of drug discovery data later. Another approach that I could try was that I could use RandomForestClassifier() instead on max phase prediction of these small molecules, rather than making pKi value predictions. This might involve re-labelling the max phases for these compounds into a binary or class labels, then I could use the imbalance-learn package to try and alleviate the problem with imbalanced datasets. Nevertheless, I had some fun working on this post and learnt a lot while doing it, and I hope some of the readers might find this post helpful or informative at least.


Acknowledgement

I’d like to thank all the authors, developers and contributors who worked towards all of the open-source packages or libraries used in this post. I’d also like to thank all of the other senior cheminformatics and ML practitioners who were sharing their work and knowledge online.

Back to top

References

Breiman, Leo. 1998. “Arcing Classifier (with Discussion and a Rejoinder by the Author).” The Annals of Statistics 26 (3). https://doi.org/10.1214/aos/1024691079.
———. 2001. “Random Forests.” Machine Learning 45 (1): 5–32. https://doi.org/10.1023/a:1010933404324.
Bruce, Peter, Andrew Bruce, and Peter Gedeck. 2020. Practical Statistics for Data Scientists. https://www.oreilly.com/library/view/practical-statistics-for/9781492072935/.
Chawla, N. V., K. W. Bowyer, L. O. Hall, and W. P. Kegelmeyer. 2002. “SMOTE: Synthetic Minority over-Sampling Technique.” Journal of Artificial Intelligence Research 16 (June): 321–57. https://doi.org/10.1613/jair.953.
Kunz, Nicholas. 2020. “SMOGN: Synthetic Minority over-Sampling Technique for Regression with Gaussian Noise.” PyPI. https://pypi.org/project/smogn/.
Lundberg, Scott M., Gabriel Erion, Hugh Chen, Alex DeGrave, Jordan M. Prutkin, Bala Nair, Ronit Katz, Jonathan Himmelfarb, Nisha Bansal, and Su-In Lee. 2020. “From Local Explanations to Global Understanding with Explainable AI for Trees.” Nature Machine Intelligence 2 (1): 2522–5839.
Shapley, Lloyd S et al. 1953. “A Value for n-Person Games.”
Torgo, Luís, Rita P. Ribeiro, Bernhard Pfahringer, and Paula Branco. 2013. “SMOTE for Regression.” In, 378–89. Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-642-40669-0_33.
Wu, Wenglei, Nicholas Kunz, and Paula Branco. 2022. “ImbalancedLearningRegression-a Python Package to Tackle the Imbalanced Regression Problem.” In Joint European Conference on Machine Learning and Knowledge Discovery in Databases, 645–48. Springer.
]]> Machine learning projects Tree models Pandas Scikit-learn ChEMBL database Python https://jhylin.github.io/Data_in_life_blog/posts/17_ML2-2_Random_forest/1_random_forest.html Tue, 21 Nov 2023 11:00:00 GMT Decision tree Jennifer HY Lin https://jhylin.github.io/Data_in_life_blog/posts/16_ML2-1_Decision_tree/1_data_col_prep.html
Series overview

  • Post 1 (this post) - data collection from ChEMBL database using web resource client in Python, with initial data preprocessing

  • Post 2 - more data preprocessing and transformation to reach the final dataset prior to model building

  • Post 3 - estimating experimental errors and building decision tree model using scikit-learn


Introduction

I’ve now come to a stage to do some more machine learning (ML) work after reading a few peer-reviewed papers about ML and drug discovery. It seemed that traditional ML methods were still indispensible performance-wise, and when used in combination with deep learning neural networks, they tend to increase prediction accuracy more. I also haven’t ventured into the practicality and usefulness of large language models in drug discovery yet, but I’m aware work in this area has been started. However, comments from experienced seniors did mention that they are still very much novel and therefore may not be as useful yet. Although by the speed of how things evolve in the so-called “AI” field, this possibly may change very soon. Also from what I can imagine, molecular representations in texts or strings are not quite the same as natural human language texts, since there are a lot of other chemistry-specific features to consider, e.g. chiralities, aromaticities and so on. Because of this, I’m sticking with learning to walk first by trying to cover conventional ML methods in a more thorough way, before trying to run in the deep learning zone.

So this leads to this series of posts (3 in total) about decision tree. Previously, I’ve only lightly touched on a commonly used classifier algorithm, logistic regression, as the first series in the ML realm. Reflecting back, I think I could’ve done a more thorough job during the data preparation stage. So this would be attempted this time. The data preparation used here was carried out with strong reference to the materials and methods section in this paper (Tilborg, Alenicheva, and Grisoni 2022), which was one of the papers I’ve read. There are probably other useful methods out there, but this paper made sense to me, so I’ve adopted a few of their ways of doing things during data preprocessing.


Data retrieval

This time I decided to try something new which was to use the ChEMBL web resource client to collect data (i.e. not by direct file downloads from ChEMBL website, although other useful way could be through SQL queries, which is also on my list to try later). I found this great online resource about fetching data this way from the TeachOpenCADD talktorial on compound data acquisition. The data retrieval workflow used below was mainly adapted from this talktorial with a few changes to suit the selected dataset and ML model.

The web resource client was supported by the ChEMBL group and was based on a Django QuerySet interface. Their GitHub repository might explain a bit more about it, particularly the Jupyter notebook link provided in the repository would help a lot regarding how to write code to search for specific data.

To do this, a few libraries needed to be loaded first.

# Import libraries
# Fetch data through ChEMBL web resource client
from chembl_webresource_client.new_client import new_client

# Dataframe library
import pandas as pd

# Progress bar
from tqdm import tqdm

To see what types of data were provided by ChEMBL web resource client, run the following code and refer to ChEMBL documentations to find out what data were embedded inside different data categories. Sometimes, it might not be that straight forward and some digging would be required (I went back to this step below to find the “data_validity_comment” when I was trying to do some compound sanitisations actually).

Note

The link provided above also talked about other useful techniques for data checks in the ChEMBL database - a very important step to do during data preprocessing, which was also something I was trying to cover and achieve as much as possible in this post.

available_resources = [resource for resource in dir(new_client) if not resource.startswith('_')]
print(available_resources)
['activity', 'activity_supplementary_data_by_activity', 'assay', 'assay_class', 'atc_class', 'binding_site', 'biotherapeutic', 'cell_line', 'chembl_id_lookup', 'compound_record', 'compound_structural_alert', 'description', 'document', 'document_similarity', 'drug', 'drug_indication', 'drug_warning', 'go_slim', 'image', 'mechanism', 'metabolism', 'molecule', 'molecule_form', 'official', 'organism', 'protein_classification', 'similarity', 'source', 'substructure', 'target', 'target_component', 'target_relation', 'tissue', 'xref_source']

Resource objects were created to enable API access as suggested by the talktorial.

# for targets (proteins)
targets_api = new_client.target

# for bioactivities
bioact_api = new_client.activity

# for assays
assay_api = new_client.assay

# for compounds
cpd_api = new_client.molecule

Checked object type for one of these API objects (e.g. bioactivity API object).

type(bioact_api)
chembl_webresource_client.query_set.QuerySet


Fetching target data

A protein target e.g. acetylcholinesterase was randomly chosen by using UniProt to look up the protein UniProt ID.

# Specify Uniprot ID for acetylcholinesterase
uniprot_id = "P22303"

# Get info from ChEMBL about this protein target, 
# with selected features only
targets = targets_api.get(target_components__accession = uniprot_id).only(
    "target_chembl_id",
    "organism", 
    "pref_name", 
    "target_type"
)

The query results were stored in a “targets” object, which was a QuerySet with lazy data evaluation only, meaning it would only react when there was a request for the data. Therefore, to see the results, the “targets” object was then read through Pandas DataFrame.

# Read "targets" with Pandas
targets = pd.DataFrame.from_records(targets)
targets
organism pref_name target_chembl_id target_type
0 Homo sapiens Acetylcholinesterase CHEMBL220 SINGLE PROTEIN
1 Homo sapiens Acetylcholinesterase CHEMBL220 SINGLE PROTEIN
2 Homo sapiens Cholinesterases; ACHE & BCHE CHEMBL2095233 SELECTIVITY GROUP

Selected the first protein target from this dataframe.

# Save the first protein in the dataframe
select_target = targets.iloc[0]
select_target
organism                    Homo sapiens
pref_name           Acetylcholinesterase
target_chembl_id               CHEMBL220
target_type               SINGLE PROTEIN
Name: 0, dtype: object

Then saved the selected ChEMBL ID for the first protein (to be used later).

chembl_id = select_target.target_chembl_id
# Check it's saved
print(chembl_id)
CHEMBL220


Fetching bioactivity data

Obtaining bioactivity data for the selected target.

bioact = bioact_api.filter(
    # Use the previously saved target ChEMBL ID
    target_chembl_id = chembl_id, 
    # Selecting for Ki
    standard_type = "Ki",
    # Requesting exact measurements
    relation = "=",
    # Binding data as "B"
    assay_type = "B",
).only(
    "activity_id",
    "data_validity_comment"
    "assay_chembl_id",
    "assay_description",
    "assay_type",
    "molecule_chembl_id",
    "standard_units",
    "standard_type",
    "relation",
    "standard_value",
    "target_chembl_id",
    "target_organism"
)

# Check the length and type of bioactivities object
print(len(bioact), type(bioact))
706 <class 'chembl_webresource_client.query_set.QuerySet'>

To have a quick look at the data being held inside each entry of the bioactivity dataset, e.g. for first entry.

print(len(bioact[0]), type(bioact[0]))
bioact[0]
15 <class 'dict'>
{'activity_id': 111024,
 'assay_chembl_id': 'CHEMBL641011',
 'assay_description': 'Inhibition constant determined against Acetylcholinesterase (AChE) receptor.',
 'assay_type': 'B',
 'data_validity_comment': 'Potential transcription error',
 'molecule_chembl_id': 'CHEMBL11805',
 'relation': '=',
 'standard_type': 'Ki',
 'standard_units': 'nM',
 'standard_value': '0.104',
 'target_chembl_id': 'CHEMBL220',
 'target_organism': 'Homo sapiens',
 'type': 'Ki',
 'units': 'nM',
 'value': '0.104'}

The next step might take a few minutes - downloading the QuerySet as a Pandas DataFrame.

bioact_df = pd.DataFrame.from_dict(bioact)

bioact_df.head(3)
activity_id assay_chembl_id assay_description assay_type data_validity_comment molecule_chembl_id relation standard_type standard_units standard_value target_chembl_id target_organism type units value
0 111024 CHEMBL641011 Inhibition constant determined against Acetylc... B Potential transcription error CHEMBL11805 = Ki nM 0.104 CHEMBL220 Homo sapiens Ki nM 0.104
1 118575 CHEMBL641012 Inhibitory activity against human AChE B None CHEMBL208599 = Ki nM 0.026 CHEMBL220 Homo sapiens Ki nM 0.026
2 125075 CHEMBL641011 Inhibition constant determined against Acetylc... B None CHEMBL60745 = Ki nM 1.63 CHEMBL220 Homo sapiens Ki nM 1.63

Checked total rows and columns in the bioactivities dataframe.

bioact_df.shape
(706, 15)


Preprocess bioactivity data

When I reached the second half of data preprocessing, an alarm bell went off regarding using half maximal inhibitory concentration (IC50) values in ChEMBL. I remembered reading recent blog posts by Greg Landrum about using IC50 and inhibition constant (Ki) values from ChEMBL. A useful open-access paper (Kalliokoski et al. 2013) from 2013 also looked into this issue about using mixed IC50 data in ChEMBL, and provided a thorough overview about how to deal with situations like this. There was also another paper (Kramer et al. 2012) on mixed Ki data from the same author group in 2012 that touched on similar issues.

To summarise both the paper about IC50 and blog posts mentioned above:

  • it would be the best to check the details of assays used to test the compounds to ensure they were aligned and not extremely heterogeneous, since IC50 values were very assay-specific, and knowing that these values were extracted from different papers from different labs all over the world, mixing them without knowing was definitely not a good idea

  • the slightly better news was that it was more likely okay to combine Ki values for the same protein target from ChEMBL as they were found to be adding less noise to the data (however ideally similar data caution should also apply)

  • it was also possible to mix Ki values with IC50 values, but the data would need to be corrected via using a conversion factor of 2.0 to convert Ki values to IC50 values (note: I also wondered if this needed to be re-looked again since this paper was published 10 years ago…)

Because of this, I decided to stick with Ki values only for now before adding more complexities as I wasn’t entirely confident about mixing IC50 values with Ki values yet. Firstly, I checked for all types of units being used in bioact_df. There were numerous different units and formats, which meant they would need to be converted to nanomolar (nM) units first.

bioact_df["units"].unique()
array(['nM', 'M', 'uM', None, 'pM', "10'-9M", "10'-3M", "10'-6M",
       "10'-10M", '/min/M', "10'5/M/min", "10'2/M/min", "10'3/M/min",
       "10'8/M/min", "10'7/M/min", 'microM/L', 'umol/L', 'mM',
       "10'4/M/min", "10'6/M/min", 'mM/min', '10^8M'], dtype=object)

Checking again that I’ve fetched Ki values only.

bioact_df["standard_type"].unique()
array(['Ki'], dtype=object)

It looked like there were duplicates of columns on units and values, so the “units” and “value” columns were removed and “standard_units” and “standard_value” columns were kept instead. Also, “type” column was dropped as there were already a “standard_type” column.

Note

Differences between “type” and “standard_type” columns were mentioned by this ChEMBL blog post.

bioact_df.drop(["units", "value", "type"], axis = 1, inplace = True)
# Re-check df
bioact_df.head(3)
activity_id assay_chembl_id assay_description assay_type data_validity_comment molecule_chembl_id relation standard_type standard_units standard_value target_chembl_id target_organism
0 111024 CHEMBL641011 Inhibition constant determined against Acetylc... B Potential transcription error CHEMBL11805 = Ki nM 0.104 CHEMBL220 Homo sapiens
1 118575 CHEMBL641012 Inhibitory activity against human AChE B None CHEMBL208599 = Ki nM 0.026 CHEMBL220 Homo sapiens
2 125075 CHEMBL641011 Inhibition constant determined against Acetylc... B None CHEMBL60745 = Ki nM 1.63 CHEMBL220 Homo sapiens 
bioact_df.dtypes
activity_id               int64
assay_chembl_id          object
assay_description        object
assay_type               object
data_validity_comment    object
molecule_chembl_id       object
relation                 object
standard_type            object
standard_units           object
standard_value           object
target_chembl_id         object
target_organism          object
dtype: object

The column of “standard_value” was converted from “object” to “float64” so we could use the Ki values for calculations later.

bioact_df = bioact_df.astype({"standard_value": "float64"})
# Check column data types again
bioact_df.dtypes
activity_id                int64
assay_chembl_id           object
assay_description         object
assay_type                object
data_validity_comment     object
molecule_chembl_id        object
relation                  object
standard_type             object
standard_units            object
standard_value           float64
target_chembl_id          object
target_organism           object
dtype: object

Then the next step was taking care of any missing entries by removing them in the first place. I excluded “data_validity_comment” column here as this was required to check if there were any unusual activity data e.g. excessively low or high Ki values. A lot of the compounds in this column probably had empty cells or “None”, which ensured that there were no particular alarm bells to the extracted bioactivity data.

bioact_df.dropna(subset = ["activity_id", "assay_chembl_id", "assay_description", "assay_type", "molecule_chembl_id", "relation",  "standard_type", "standard_units", "standard_value", "target_chembl_id", "target_organism"], axis = 0, how = "any", inplace = True)
# Check number of rows and columns again (in this case, there appeared to be no change for rows)
bioact_df.shape
(706, 12)

Since all unique units inside the “units” and “values” columns were checked previously, I’d done the same for the “standard_units” column to see the ones recorded in it.

bioact_df["standard_units"].unique()
array(['nM', '/min/M', "10'5/M/min", "10'2/M/min", "10'3/M/min",
       "10'8/M/min", "10'7/M/min", "10'4/M/min", "10'6/M/min", 'mM/min',
       '10^8M'], dtype=object)

There were a mixture of different units.

# Check for number of non-nM units
bioact_df[bioact_df["standard_units"] != "nM"].shape[0]
61

There appeared to be 61 non-nM values inside the fetched bioactivity data.

bioact_df = bioact_df[bioact_df["standard_units"] == "nM"]

I then narrowed the results to only “nM” and checked the dataframe again to see what units were left now.

# Check there were only nM
bioact_df["standard_units"].unique()
array(['nM'], dtype=object)

So the filtering worked and the number of rows and columns were reduced.

# Check df rows & columns
bioact_df.shape
(645, 12)

Next part would be to remove all the duplicates in the dataframe, especially when there were duplicate tests for the same compound.

bioact_df.drop_duplicates("molecule_chembl_id", keep = "first", inplace = True)

Renamed the “standard_value” and “standard_units” columns to “Ki” and “units” respectively.

bioact_df.rename(
    columns = {
        "standard_value": "Ki",
        "standard_units": "units"
    }, inplace = True
)

# Check df to ensure name change
bioact_df.head(3)
activity_id assay_chembl_id assay_description assay_type data_validity_comment molecule_chembl_id relation standard_type units Ki target_chembl_id target_organism
0 111024 CHEMBL641011 Inhibition constant determined against Acetylc... B Potential transcription error CHEMBL11805 = Ki nM 0.104 CHEMBL220 Homo sapiens
1 118575 CHEMBL641012 Inhibitory activity against human AChE B None CHEMBL208599 = Ki nM 0.026 CHEMBL220 Homo sapiens
2 125075 CHEMBL641011 Inhibition constant determined against Acetylc... B None CHEMBL60745 = Ki nM 1.630 CHEMBL220 Homo sapiens

Lastly, the index of the dataframe was reset.

bioact_df.reset_index(drop = True, inplace = True)
bioact_df.head(3)
activity_id assay_chembl_id assay_description assay_type data_validity_comment molecule_chembl_id relation standard_type units Ki target_chembl_id target_organism
0 111024 CHEMBL641011 Inhibition constant determined against Acetylc... B Potential transcription error CHEMBL11805 = Ki nM 0.104 CHEMBL220 Homo sapiens
1 118575 CHEMBL641012 Inhibitory activity against human AChE B None CHEMBL208599 = Ki nM 0.026 CHEMBL220 Homo sapiens
2 125075 CHEMBL641011 Inhibition constant determined against Acetylc... B None CHEMBL60745 = Ki nM 1.630 CHEMBL220 Homo sapiens

One final check on the number of columns and rows after preprocessing the bioactivity dataframe.

bioact_df.shape
(540, 12)

There were a total of 12 columns with 540 rows of data left in the bioactivity dataframe.


Fetching assay data

The assay data was added after I went through the rest of the data preprocessing and also after remembering to check on the confidence scores for assays used in the final data collected (to somewhat assess assay-to-target relationships). This link from ChEMBL explained what the confidence score meant.

assays = assay_api.filter(
    # Use the previously saved target ChEMBL ID
    target_chembl_id = chembl_id, 
    # Binding assays only as before
    assay_type = "B"
).only(
    "assay_chembl_id",
    "confidence_score"
)

Placing the fetched assay data into a Pandas DataFrame.

assays_df = pd.DataFrame.from_dict(assays)

print(assays_df.shape)
assays_df.head(3)
(2044, 2)
assay_chembl_id confidence_score
0 CHEMBL634034 8
1 CHEMBL642512 8
2 CHEMBL642513 8 
assays_df.describe()
confidence_score
count 2044.000000
mean 8.778865
std 0.415113
min 8.000000
25% 9.000000
50% 9.000000
75% 9.000000
max 9.000000

It looked like the lowest confidence score for this particular protein target in binding assays was at 8, with others sitting at 9 (the highest). There were 452 assays with confidence score of 8.

# Some had score of 8 - find out which ones
assays_df[assays_df["confidence_score"] == 8]
assay_chembl_id confidence_score
0 CHEMBL634034 8
1 CHEMBL642512 8
2 CHEMBL642513 8
3 CHEMBL642514 8
4 CHEMBL642515 8
... ... ...
1141 CHEMBL3887379 8
1142 CHEMBL3887855 8
1143 CHEMBL3887947 8
1144 CHEMBL3888161 8
1874 CHEMBL5058677 8

452 rows × 2 columns


Combining bioactivity & assay data

The key was to combine the bioactivity and assay data along the “assay_chembl_id” column.

bioact_assay_df = pd.merge(
    bioact_df[["assay_chembl_id", "molecule_chembl_id", "Ki", "units", "data_validity_comment"]],
    assays_df,
    on = "assay_chembl_id",
)
print(bioact_assay_df.shape)
bioact_assay_df.head(3)
(540, 6)
assay_chembl_id molecule_chembl_id Ki units data_validity_comment confidence_score
0 CHEMBL641011 CHEMBL11805 0.104 nM Potential transcription error 8
1 CHEMBL641011 CHEMBL60745 1.630 nM None 8
2 CHEMBL641012 CHEMBL208599 0.026 nM None 8

I actually came back to this step to relax the confidence score limit to include all the 8s as well as the 9s (otherwise previously I tried only using assays with score of 9), so that donepezil and galantamine could be included in the dataset as well (the purpose of this would be clearer in post 3 when building the model).


Fetching compound data

While having identified the protein target, obtained the bioactivity data, and also the assay data, this next step was to fetch the compound data. This could be done by having the ChEMBL IDs available in the bioactivity dataset.

cpds = cpd_api.filter(
    molecule_chembl_id__in = list(bioact_df["molecule_chembl_id"])
).only(
    "molecule_chembl_id",
    "molecule_structures",
    "max_phase"
)

Here, the same step was applied where the compound QuerySet object was converted into a Pandas dataframe. However, the compound data extracted here might take longer than the bioactivity one. One way to monitor progress was through using tqdm package.

compds = list(tqdm(cpds))
  0%|          | 0/540 [00:00<?, ?it/s]
 93%|█████████▎| 501/540 [00:00<00:00, 4845.36it/s]
100%|██████████| 540/540 [00:00<00:00, 5020.27it/s]

Converting retrieved compound QuerySet into a Pandas DataFrame.

cpds_df = pd.DataFrame.from_records(compds)
print(cpds_df.shape)
cpds_df.head(3)
(540, 3)
max_phase molecule_chembl_id molecule_structures
0 None CHEMBL28 {'canonical_smiles': 'O=c1cc(-c2ccc(O)cc2)oc2c...
1 3.0 CHEMBL50 {'canonical_smiles': 'O=c1c(O)c(-c2ccc(O)c(O)c...
2 None CHEMBL8320 {'canonical_smiles': 'O=C1C=CC(=O)C=C1', 'molf...


Preprocess compound data

Removing any missing entries in the compound data (excluding the “max_phase” column as it was needed during the model training/testing part in post 3 - note: “None” entries meant they were preclinical molecules so not assigned with a max phase yet).

cpds_df.dropna(subset = ["molecule_chembl_id", "molecule_structures"], axis = 0, how = "any", inplace = True)

# Check columns & rows in df
cpds_df.shape
(540, 3)

Removing any duplicates in the compound data.

cpds_df.drop_duplicates("molecule_chembl_id", keep = "first", inplace = True)

# Check columns & rows again
cpds_df.shape
(540, 3)

Ideally, only the compounds with canonical SMILES would be kept. Checking for the types of molecular representations used in the “molecule_structures” column of the compound dataset.

# Randomly choosing the 2nd entry as example
cpds_df.iloc[1].molecule_structures.keys()
dict_keys(['canonical_smiles', 'molfile', 'standard_inchi', 'standard_inchi_key'])

There were 4 types: “canonical_smiles”, “molfile”, “standard_inchi” and “standard_inchi_key”.

# Create an empty list to store the canonical smiles
can_smiles = []

# Create a for loop to loop over each row of data, 
# searching for only canonical_smiles to append to the created list
for i, cpd in cpds_df.iterrows():
    try:
        can_smiles.append(cpd["molecule_structures"]["canonical_smiles"])
    except KeyError:
        can_smiles.append(None)

# Create a new df column with name as "smiles", 
# which will store all the canonical smiles collected from the list above
cpds_df["smiles"] = can_smiles

Check the compound dataframe quickly to see if a new column for SMILES has been created.

cpds_df.head(3)
max_phase molecule_chembl_id molecule_structures smiles
0 None CHEMBL28 {'canonical_smiles': 'O=c1cc(-c2ccc(O)cc2)oc2c... O=c1cc(-c2ccc(O)cc2)oc2cc(O)cc(O)c12
1 3.0 CHEMBL50 {'canonical_smiles': 'O=c1c(O)c(-c2ccc(O)c(O)c... O=c1c(O)c(-c2ccc(O)c(O)c2)oc2cc(O)cc(O)c12
2 None CHEMBL8320 {'canonical_smiles': 'O=C1C=CC(=O)C=C1', 'molf... O=C1C=CC(=O)C=C1

Once confirmed, the old “molecule_structures” column was then removed.

cpds_df.drop("molecule_structures", axis = 1, inplace = True)

Finally, adding another step to ensure all missing entries or entries without canonical SMILES strings were removed from the compound dataset.

cpds_df.dropna(subset = ["smiles"], axis = 0, how = "any", inplace = True)

print(cpds_df.shape)
(540, 3)

Final look at the compound dataset, which should only include max phase, compound ChEMBL IDs and SMILES columns.

cpds_df.head(3)
max_phase molecule_chembl_id smiles
0 None CHEMBL28 O=c1cc(-c2ccc(O)cc2)oc2cc(O)cc(O)c12
1 3.0 CHEMBL50 O=c1c(O)c(-c2ccc(O)c(O)c2)oc2cc(O)cc(O)c12
2 None CHEMBL8320 O=C1C=CC(=O)C=C1


Combining bioactivity and compound data

To combine both datasets, the key was to look for common column (similar to a SQL “join” query) between the two datasets.

Listing all the column names for both datasets.

bioact_assay_df.columns
Index(['assay_chembl_id', 'molecule_chembl_id', 'Ki', 'units',
       'data_validity_comment', 'confidence_score'],
      dtype='object')
cpds_df.columns
Index(['max_phase', 'molecule_chembl_id', 'smiles'], dtype='object')

Clearly, the column that existed in both dataframes was the “molecule_chembl_id” column.

The next step was to combine or merge both datasets.

# Create a final dataframe that will contain both bioactivity and compound data
dtree_df = pd.merge(
    bioact_assay_df[["molecule_chembl_id","Ki", "units", "data_validity_comment"]],
    cpds_df,
    on = "molecule_chembl_id",
)

dtree_df.head(3)
molecule_chembl_id Ki units data_validity_comment max_phase smiles
0 CHEMBL11805 0.104 nM Potential transcription error None COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)...
1 CHEMBL60745 1.630 nM None None CC[N+](C)(C)c1cccc(O)c1.[Br-]
2 CHEMBL208599 0.026 nM None None CCC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2

Shape of the final dataframe was checked.

print(dtree_df.shape)
(540, 6)

Saving a copy of the merged dataframe for now to avoid re-running the previous code repeatedly, and also to be ready for second-half of the data preprocessing work, which will be in post 2.

dtree_df.to_csv("ache_chembl.csv")
Back to top

References

Kalliokoski, Tuomo, Christian Kramer, Anna Vulpetti, and Peter Gedeck. 2013. “Comparability of Mixed IC50 Data A Statistical Analysis.” Edited by Andrea Cavalli. PLoS ONE 8 (4): e61007. https://doi.org/10.1371/journal.pone.0061007.
Kramer, Christian, Tuomo Kalliokoski, Peter Gedeck, and Anna Vulpetti. 2012. “The Experimental Uncertainty of Heterogeneous Public Ki Data.” Journal of Medicinal Chemistry 55 (11): 5165–73. https://doi.org/10.1021/jm300131x.
Tilborg, Derek van, Alisa Alenicheva, and Francesca Grisoni. 2022. “Exposing the Limitations of Molecular Machine Learning with Activity Cliffs.” Journal of Chemical Information and Modeling 62 (23): 5938–51. https://doi.org/10.1021/acs.jcim.2c01073.
]]> Machine learning projects Tree models Data preprocessing Pandas ChEMBL database Python https://jhylin.github.io/Data_in_life_blog/posts/16_ML2-1_Decision_tree/1_data_col_prep.html Mon, 18 Sep 2023 12:00:00 GMT Decision tree Jennifer HY Lin https://jhylin.github.io/Data_in_life_blog/posts/16_ML2-1_Decision_tree/2_data_prep_tran.html
Data source

The data used in this post 2 for data preprocessing was extracted from ChEMBL database by using ChEMBL web resource client in Python. The details of all the steps taken to reach the final .csv file could be seen in post 1.


Checklist for preprocessing ChEMBL compound data

Below was a checklist summary for post 1 and post 2 (current post), and was highly inspired by this journal paper (Tilborg, Alenicheva, and Grisoni 2022) and also ChEMBL’s FAQ on “Assay and Activity Questions”.

Note: not an exhaustive list, only a suggestion from my experience working on this series, may need to tailor to different scenarios

For molecular data containing chemical compounds, check for:

  • duplicates
  • missing values
  • salts or mixture

Check the consistency of structural annotations:

  • molecular validity
  • molecular sanity
  • charge standardisation
  • stereochemistry

Check the reliability of reported experimental values (e.g. activity values like IC50, Ki, EC50 etc.):

  • annotated validity (data_validity_comment)
  • presence of outliers
  • confidence score (assays)
  • standard deviation of multiple entries (if applicable)


Import libraries
# Import all libraries used
import pandas as pd
import math
from rdkit.Chem import Descriptors
import datamol as dm
# tqdm library used in datamol's batch descriptor code
from tqdm import tqdm
import mols2grid


Re-import saved data

Re-imported the partly preprocessed data from the earlier post.

dtree_df = pd.read_csv("ache_chembl.csv")
dtree_df.head(3)
Unnamed: 0 molecule_chembl_id Ki units data_validity_comment max_phase smiles
0 0 CHEMBL11805 0.104 nM Potential transcription error NaN COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)...
1 1 CHEMBL60745 1.630 nM NaN NaN CC[N+](C)(C)c1cccc(O)c1.[Br-]
2 2 CHEMBL208599 0.026 nM NaN NaN CCC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2

There was an extra index column (named “Unnamed: 0”) here, which was likely inherited from how the .csv file was saved with the index already in place from part 1, so this column was dropped for now.

dtree_df = dtree_df.drop("Unnamed: 0", axis = 1)
dtree_df.head(3)
molecule_chembl_id Ki units data_validity_comment max_phase smiles
0 CHEMBL11805 0.104 nM Potential transcription error NaN COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)...
1 CHEMBL60745 1.630 nM NaN NaN CC[N+](C)(C)c1cccc(O)c1.[Br-]
2 CHEMBL208599 0.026 nM NaN NaN CCC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2


Calculate pKi

The distribution of Ki values were shown below via a simple statistical summary.

dtree_df["Ki"].describe()
count    5.400000e+02
mean     2.544039e+05
std      4.103437e+06
min      1.700000e-03
25%      2.437500e+01
50%      1.995000e+02
75%      3.100000e+03
max      9.496300e+07
Name: Ki, dtype: float64

From the above quick statistical summary and also the code below to find the minimum Ki value, it confirmed there were no zero Ki values recorded.

dtree_df["Ki"].min()
0.0017

Now the part about converting the Ki values to pKi values, which were the negative logs of Ki in molar units (a PubChem example might help to explain it a little). The key to understand pKi here was to treat pKi similarly to how we normally understand pH for our acids and bases. The formula to convert Ki to pKi for nanomolar (nM) units was:

Set up a small function to do the conversion.

def calc_pKi(Ki):
    pKi_value = 9 - math.log10(Ki)
    return pKi_value

Applying the calc_pKi function to convert all rows of the compound dataset for the “Ki” column.

# Create a new column for pKi
# Apply calc_pKi function to data in Ki column
dtree_df["pKi"] = dtree_df.apply(lambda x: calc_pKi(x.Ki), axis = 1)

The dataframe would now look like this, with a new pKi column (scroll to the very right to see it).

dtree_df.head(3)
molecule_chembl_id Ki units data_validity_comment max_phase smiles pKi
0 CHEMBL11805 0.104 nM Potential transcription error NaN COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)... 9.982967
1 CHEMBL60745 1.630 nM NaN NaN CC[N+](C)(C)c1cccc(O)c1.[Br-] 8.787812
2 CHEMBL208599 0.026 nM NaN NaN CCC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2 10.585027


Plan other data preprocessing steps

For a decision tree model, a few more molecular descriptors were most likely needed rather than only Ki or pKi and SMILES, since I’ve now arrived at the step of planning other preprocessing steps. One way to do this could be through computations based on canonical SMILES of compounds by using RDKit, which would give the RDKit 2D descriptors. In this single tree model, I decided to stick with only RDKit 2D descriptors for now, before adding on fingerprints (as a side note: I have very lightly touched on generating fingerprints in this earlier post - “Molecular similarities in selected COVID-19 antivirals” in the subsection on “Fingerprint generator”).

At this stage, a compound sanitisation step should also be applied to the compound column before starting any calculations to rule out compounds with questionable chemical validities. RDKit or Datamol (a Python wrapper library built based on RDKit) was also capable of doing this.

I’ve added a quick step here to convert the data types of “smiles” and “data_validity_comment” columns to string (in case of running into problems later).

dtree_df = dtree_df.astype({"smiles": "string", "data_validity_comment": "string"})
dtree_df.dtypes
molecule_chembl_id        object
Ki                       float64
units                     object
data_validity_comment     string
max_phase                float64
smiles                    string
pKi                      float64
dtype: object


Check data validity

Also, before jumping straight to compound sanitisation, I needed to check the “data_validity_comment” column.

dtree_df["data_validity_comment"].unique()
<StringArray>
['Potential transcription error', <NA>, 'Outside typical range']
Length: 3, dtype: string

There were 3 different types of data validity comments here, which were “Potential transcription error”, “NaN” and “Outside typical range”. So, this meant compounds with comments as “Potential transcription error” and “Outside typical range” should be addressed first.

# Find out number of compounds with "outside typical range" as data validity comment
dtree_df_err = dtree_df[dtree_df["data_validity_comment"] == "Outside typical range"]
print(dtree_df_err.shape)
dtree_df_err.head()
(58, 7)
molecule_chembl_id Ki units data_validity_comment max_phase smiles pKi
111 CHEMBL225198 0.0090 nM Outside typical range NaN O=C(CCc1c[nH]c2ccccc12)NCCCCCCCNc1c2c(nc3cc(Cl... 11.045757
114 CHEMBL225021 0.0017 nM Outside typical range NaN O=C(CCCc1c[nH]c2ccccc12)NCCCCCNc1c2c(nc3cc(Cl)... 11.769551
118 CHEMBL402976 313700.0000 nM Outside typical range NaN CN(C)CCOC(=O)Nc1ccncc1 3.503485
119 CHEMBL537454 140200.0000 nM Outside typical range NaN CN(C)CCOC(=O)Nc1cc(Cl)nc(Cl)c1.Cl 3.853252
120 CHEMBL3216883 316400.0000 nM Outside typical range NaN CN(C)CCOC(=O)Nc1ccncc1Br.Cl.Cl 3.499764

There were a total of 58 compounds with Ki outside typical range.

dtree_df_err2 = dtree_df[dtree_df["data_validity_comment"] == "Potential transcription error"]
dtree_df_err2
molecule_chembl_id Ki units data_validity_comment max_phase smiles pKi
0 CHEMBL11805 0.104 nM Potential transcription error NaN COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)... 9.982967

With the other comment for potential transciption error, there seemed to be only one compound here.

These compounds with questionable Ki values were removed, as they could be potential sources of errors in ML models later on (error trickling effect). One of the ways to filter out data was to fill the empty cells within the “data_validity_comment” column first, so those ones to be kept could be selected.

# Fill "NaN" entries with an actual name e.g. none
dtree_df["data_validity_comment"].fillna("none", inplace=True)
dtree_df.head()
molecule_chembl_id Ki units data_validity_comment max_phase smiles pKi
0 CHEMBL11805 0.104 nM Potential transcription error NaN COc1ccccc1CN(C)CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)... 9.982967
1 CHEMBL60745 1.630 nM none NaN CC[N+](C)(C)c1cccc(O)c1.[Br-] 8.787812
2 CHEMBL208599 0.026 nM none NaN CCC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2 10.585027
3 CHEMBL95 151.000 nM none 4.0 Nc1c2c(nc3ccccc13)CCCC2 6.821023
4 CHEMBL173309 12.200 nM none NaN CCN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)C... 7.913640

Filtered out only the compounds with nil data validity comments.

#dtree_df["data_validity_comment"].unique()
dtree_df = dtree_df[dtree_df["data_validity_comment"] == "none"]

Checking the dtree_df dataframe again and also whether if only the compounds with “none” labelled for “data_validity_comment” column were kept.

print(dtree_df.shape)
dtree_df["data_validity_comment"].unique()
(481, 7)
<StringArray>
['none']
Length: 1, dtype: string


Sanitise compounds

This preprocessing molecules tutorial and reference links provided by Datamol were very informative, and the preprocess function code by Datamol was used below. Each step of fix_mol(), sanitize_mol() and standardize_mol() was explained in this tutorial. I think the key was to select preprocessing options required to fit the purpose of the ML models, and the more experiences in doing this, the more likely it will help with the preprocessing step.

# _preprocess function to sanitise compounds - adapted from datamol.io

smiles_column = "smiles"

dm.disable_rdkit_log()

def _preprocess(row):
    # Convert each compound to a RDKit molecule in the smiles column
    mol = dm.to_mol(row[smiles_column], ordered=True)
    # Fix common errors in the molecules
    mol = dm.fix_mol(mol)
    # Sanitise the molecules 
    mol = dm.sanitize_mol(mol, sanifix=True, charge_neutral=False)
    # Standardise the molecules
    mol = dm.standardize_mol(
        mol,
        # Switch on to disconnect metal ions
        disconnect_metals=True,
        normalize=True,
        reionize=True,
        # Switch on "uncharge" to neutralise charges
        uncharge=True,
        # Taking care of stereochemistries of compounds
        stereo=True,
    )

    # Added a new column below for RDKit molecules
    row["rdkit_mol"] = dm.to_mol(mol)
    row["standard_smiles"] = dm.standardize_smiles(dm.to_smiles(mol))
    row["selfies"] = dm.to_selfies(mol)
    row["inchi"] = dm.to_inchi(mol)
    row["inchikey"] = dm.to_inchikey(mol)
    return row

Then the compound sanitisation function was applied to the dtree_df.

dtree_san_df = dtree_df.apply(_preprocess, axis = 1)
dtree_san_df.head()
molecule_chembl_id Ki units data_validity_comment max_phase smiles pKi rdkit_mol standard_smiles selfies inchi inchikey
1 CHEMBL60745 1.630 nM none NaN CC[N+](C)(C)c1cccc(O)c1.[Br-] 8.787812 <rdkit.Chem.rdchem.Mol object at 0x120080f90> CC[N+](C)(C)c1cccc(O)c1.[Br-] [C][C][N+1][Branch1][C][C][Branch1][C][C][C][=... InChI=1S/C10H15NO.BrH/c1-4-11(2,3)9-6-5-7-10(1... CAEPIUXAUPYIIJ-UHFFFAOYSA-N
2 CHEMBL208599 0.026 nM none NaN CCC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2 10.585027 <rdkit.Chem.rdchem.Mol object at 0x120081cb0> CCC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2 [C][C][C][=C][C][C][C][=N][C][=C][C][Branch1][... InChI=1S/C18H19ClN2/c1-2-10-5-11-7-12(6-10)17-... QTPHSDHUHXUYFE-KIYNQFGBSA-N
3 CHEMBL95 151.000 nM none 4.0 Nc1c2c(nc3ccccc13)CCCC2 6.821023 <rdkit.Chem.rdchem.Mol object at 0x120081700> Nc1c2c(nc3ccccc13)CCCC2 [N][C][=C][C][=Branch1][N][=N][C][=C][C][=C][C... InChI=1S/C13H14N2/c14-13-9-5-1-3-7-11(9)15-12-... YLJREFDVOIBQDA-UHFFFAOYSA-N
4 CHEMBL173309 12.200 nM none NaN CCN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)C... 7.913640 <rdkit.Chem.rdchem.Mol object at 0x1200af450> CCN(CCCCCC(=O)N(C)CCCCCCCCN(C)C(=O)CCCCCN(CC)C... [C][C][N][Branch2][Branch1][Ring1][C][C][C][C]... InChI=1S/C42H70N4O4/c1-7-45(35-37-25-17-19-27-... VJXLWYGKZGTXAF-UHFFFAOYSA-N
5 CHEMBL1128 200.000 nM none 4.0 CC[N+](C)(C)c1cccc(O)c1.[Cl-] 6.698970 <rdkit.Chem.rdchem.Mol object at 0x1200af3e0> CC[N+](C)(C)c1cccc(O)c1.[Cl-] [C][C][N+1][Branch1][C][C][Branch1][C][C][C][=... InChI=1S/C10H15NO.ClH/c1-4-11(2,3)9-6-5-7-10(1... BXKDSDJJOVIHMX-UHFFFAOYSA-N

If the dataset required for sanitisation is large, Datamol has suggested using their example code to add parallelisation as shown below.

```{python}
# Code adapted from: https://docs.datamol.io/stable/tutorials/Preprocessing.html#references
data_clean = dm.parallelized(
    _preprocess, 
    data.iterrows(), 
    arg_type="args", 
    progress=True, 
    total=len(data)
    )
data_clean = pd.DataFrame(data_clean)
```
dtree_san_df.shape
(481, 12)

In this case, I tried using the preprocessing function without adding parallelisation, the whole process wasn’t very long (since I had a small dataset), and was done within a minute or so.

Also, as a sanity check on the sanitised compounds in dtree_san_df, I just wanted to see if I could display all compounds in this dataframe as 2D images. I also had a look through each page just to see if there were any odd bonds or anything strange in general.

# Create a list to store all cpds in dtree_san_df
mol_list = dtree_san_df["rdkit_mol"]
# Convert to list
mol_list = list(mol_list)
# Check data type
type(mol_list)
# Show 2D compound structures in grids
mols2grid.display(mol_list)


Detect outliers

Plotting a histogram to see the distribution of pKi values first.

dtree_san_df.hist(column = "pKi")
array([[<AxesSubplot: title={'center': 'pKi'}>]], dtype=object)

I read a bit about Dixon’s Q test and realised that there were a few required assumptions prior to using this test, and the current dataset used here (dtree_san_df) might not fit the requirements, which were:

  • normally distributed data
  • a small sample size e.g. between 3 and 10, which was originally stated in this paper (Dean and Dixon 1951).

I’ve also decided that rather than showing Python code for Dixon’s Q test myself, I would attach a few examples from others instead, for example, Q test from Plotly and Dixon’s Q test for outlier identification – a questionable practice, since this dataset here wasn’t quite normally distributed as shown from the histogram above.

dtree_san_df.boxplot(column = "pKi")

# the boxplot version below shows a blank background
# rather than above version with horizontal grid lines
#dtree_san_df.plot.box(column = "pKi")
<AxesSubplot: >

I also used Pandas’ built-in boxplot in addition to the histogram to show the possible outliers within the pKi values. Clearly, the outliers for pKi values appeared to be above 10. I also didn’t remove these outliers completely due to the dataset itself wasn’t quite in a Gaussian distribution (they might not be true outliers).


Calculate RDKit 2D molecular descriptors

I’ve explored a few different ways to compute molecular descriptors, essentially RDKit was used as the main library to do this (there might be other options via other programming languages, but I was only exploring RDKit-based methods via Python for now). A blog post I’ve come across on calculating RDKit 2D molecular descriptors has explained it well, it gave details about how to bundle the functions together in a class (the idea of building a small library yourself to be used in projects was quite handy). I’ve also read RDKit’s documentations and also the ones from Datamol. So rather than re-inventing the wheels of all the RDKit code, I’ve opted to use only a small chunk of RDKit code as a demonstration, then followed by Datamol’s version to compute the 2D descriptors, since there were already a few really well-explained blog posts about this. One of the examples was this useful descriptor calculation tutorial by Greg Landrum.


RDKit code

With the lastest format of the dtree_san_df, it already included a RDKit molecule column (named “rdkit_mol”), so this meant I could go ahead with the calculations. So here I used RDKit’s Descriptors.CalcMolDescriptors() to calculate the 2D descriptors - note: there might be more code variations depending on needs, this was just a small example.

# Run descriptor calculations on mol_list (created earlier)
# and save as a new list
mol_rdkit_ls = [Descriptors.CalcMolDescriptors(mol) for mol in mol_list]

# Convert the list into a dataframe
df_rdkit_2d = pd.DataFrame(mol_rdkit_ls)
print(df_rdkit_2d.shape)
df_rdkit_2d.head(3)
(481, 209)
MaxAbsEStateIndex MaxEStateIndex MinAbsEStateIndex MinEStateIndex qed MolWt HeavyAtomMolWt ExactMolWt NumValenceElectrons NumRadicalElectrons ... fr_sulfide fr_sulfonamd fr_sulfone fr_term_acetylene fr_tetrazole fr_thiazole fr_thiocyan fr_thiophene fr_unbrch_alkane fr_urea
0 9.261910 9.261910 0.000000 0.000000 0.662462 246.148 230.020 245.041526 74 0 ... 0 0 0 0 0 0 0 0 0 0
1 6.509708 6.509708 0.547480 0.547480 0.763869 298.817 279.665 298.123676 108 0 ... 0 0 0 0 0 0 0 0 0 0
2 6.199769 6.199769 0.953981 0.953981 0.706488 198.269 184.157 198.115698 76 0 ... 0 0 0 0 0 0 0 0 0 0

3 rows × 209 columns

In total, it generated 209 descriptors.


Datamol code

Then I tested Datamol’s code on this as shown below.

# Datamol's batch descriptor code for a list of compounds
dtree_san_df_dm = dm.descriptors.batch_compute_many_descriptors(mol_list)
print(dtree_san_df_dm.shape)
dtree_san_df_dm.head(3)
(481, 22)
mw fsp3 n_lipinski_hba n_lipinski_hbd n_rings n_hetero_atoms n_heavy_atoms n_rotatable_bonds n_radical_electrons tpsa ... sas n_aliphatic_carbocycles n_aliphatic_heterocyles n_aliphatic_rings n_aromatic_carbocycles n_aromatic_heterocyles n_aromatic_rings n_saturated_carbocycles n_saturated_heterocyles n_saturated_rings
0 245.041526 0.400000 2 1 1 3 13 2 0 20.23 ... 3.185866 0 0 0 1 0 1 0 0 0
1 298.123676 0.388889 2 2 4 3 21 1 0 38.91 ... 4.331775 2 0 2 1 1 2 0 0 0
2 198.115698 0.307692 2 2 3 2 15 0 0 38.91 ... 2.014719 1 0 1 1 1 2 0 0 0

3 rows × 22 columns

There were a total of 22 molecular descriptors generated, which seemed more like what I might use for the decision tree model. The limitation with this batch descriptor code was that the molecular features were pre-selected, so if other types were needed, it would be the best to go for RDKit code or look into other Datamol descriptor code that allow users to specify features. The types of descriptors were shown below.

dtree_san_df_dm.columns
Index(['mw', 'fsp3', 'n_lipinski_hba', 'n_lipinski_hbd', 'n_rings',
       'n_hetero_atoms', 'n_heavy_atoms', 'n_rotatable_bonds',
       'n_radical_electrons', 'tpsa', 'qed', 'clogp', 'sas',
       'n_aliphatic_carbocycles', 'n_aliphatic_heterocyles',
       'n_aliphatic_rings', 'n_aromatic_carbocycles', 'n_aromatic_heterocyles',
       'n_aromatic_rings', 'n_saturated_carbocycles',
       'n_saturated_heterocyles', 'n_saturated_rings'],
      dtype='object')


Combine dataframes

The trickier part for data preprocessing was actually trying to merge, join or concatenate dataframes of the preprocessed dataframe (dtree_san_df) and the dataframe from Datamol’s descriptor code (dtree_san_df_dm).

Initially, I tried using all of Pandas’ code of merge/join/concat() dataframes. They all failed to create the correct final combined dataframe with too many rows generated, with one run actually created more than 500 rows (maximum should be 481 rows). One of the possible reasons for this could be that some of the descriptors had zeros generated as results for some of the compounds, and when combining dataframes using Pandas code like the ones mentioned here, they might cause unexpected results (as suggested by Pandas, these code were not exactly equivalent to SQL joins). So I looked into different ways, and while there were no other common columns for both dataframes, the index column seemed to be the only one that correlated both.

I also found out after going back to the previous steps that when I applied the compound preprocessing function from Datamol, the index of the resultant dataframe was changed to start from 1 (rather than zero). Because of this, I tried re-setting the index of dtree_san_df first, then dropped the index column, followed by re-setting the index again to ensure it started at zero, which has worked. So now the dtree_san_df would have exactly the same index as the one for dtree_san_df_dm.

# 1st index re-set
dtree_san_df = dtree_san_df.reset_index()
# Drop the index column
dtree_san_df = dtree_san_df.drop(["index"], axis = 1)
dtree_san_df.head(3)
molecule_chembl_id Ki units data_validity_comment max_phase smiles pKi rdkit_mol standard_smiles selfies inchi inchikey
0 CHEMBL60745 1.630 nM none NaN CC[N+](C)(C)c1cccc(O)c1.[Br-] 8.787812 <rdkit.Chem.rdchem.Mol object at 0x120080f90> CC[N+](C)(C)c1cccc(O)c1.[Br-] [C][C][N+1][Branch1][C][C][Branch1][C][C][C][=... InChI=1S/C10H15NO.BrH/c1-4-11(2,3)9-6-5-7-10(1... CAEPIUXAUPYIIJ-UHFFFAOYSA-N
1 CHEMBL208599 0.026 nM none NaN CCC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2 10.585027 <rdkit.Chem.rdchem.Mol object at 0x120081cb0> CCC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2 [C][C][C][=C][C][C][C][=N][C][=C][C][Branch1][... InChI=1S/C18H19ClN2/c1-2-10-5-11-7-12(6-10)17-... QTPHSDHUHXUYFE-KIYNQFGBSA-N
2 CHEMBL95 151.000 nM none 4.0 Nc1c2c(nc3ccccc13)CCCC2 6.821023 <rdkit.Chem.rdchem.Mol object at 0x120081700> Nc1c2c(nc3ccccc13)CCCC2 [N][C][=C][C][=Branch1][N][=N][C][=C][C][=C][C... InChI=1S/C13H14N2/c14-13-9-5-1-3-7-11(9)15-12-... YLJREFDVOIBQDA-UHFFFAOYSA-N 
# 2nd index re-set
dtree_san_df = dtree_san_df.reset_index()
print(dtree_san_df.shape)
dtree_san_df.head(3)
(481, 13)
index molecule_chembl_id Ki units data_validity_comment max_phase smiles pKi rdkit_mol standard_smiles selfies inchi inchikey
0 0 CHEMBL60745 1.630 nM none NaN CC[N+](C)(C)c1cccc(O)c1.[Br-] 8.787812 <rdkit.Chem.rdchem.Mol object at 0x120080f90> CC[N+](C)(C)c1cccc(O)c1.[Br-] [C][C][N+1][Branch1][C][C][Branch1][C][C][C][=... InChI=1S/C10H15NO.BrH/c1-4-11(2,3)9-6-5-7-10(1... CAEPIUXAUPYIIJ-UHFFFAOYSA-N
1 1 CHEMBL208599 0.026 nM none NaN CCC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2 10.585027 <rdkit.Chem.rdchem.Mol object at 0x120081cb0> CCC1=CC2Cc3nc4cc(Cl)ccc4c(N)c3[C@@H](C1)C2 [C][C][C][=C][C][C][C][=N][C][=C][C][Branch1][... InChI=1S/C18H19ClN2/c1-2-10-5-11-7-12(6-10)17-... QTPHSDHUHXUYFE-KIYNQFGBSA-N
2 2 CHEMBL95 151.000 nM none 4.0 Nc1c2c(nc3ccccc13)CCCC2 6.821023 <rdkit.Chem.rdchem.Mol object at 0x120081700> Nc1c2c(nc3ccccc13)CCCC2 [N][C][=C][C][=Branch1][N][=N][C][=C][C][=C][C... InChI=1S/C13H14N2/c14-13-9-5-1-3-7-11(9)15-12-... YLJREFDVOIBQDA-UHFFFAOYSA-N

Also re-setting the index of the dtree_san_df_dm.

dtree_san_df_dm = dtree_san_df_dm.reset_index()
print(dtree_san_df_dm.shape)
dtree_san_df_dm.head(3)
(481, 23)
index mw fsp3 n_lipinski_hba n_lipinski_hbd n_rings n_hetero_atoms n_heavy_atoms n_rotatable_bonds n_radical_electrons ... sas n_aliphatic_carbocycles n_aliphatic_heterocyles n_aliphatic_rings n_aromatic_carbocycles n_aromatic_heterocyles n_aromatic_rings n_saturated_carbocycles n_saturated_heterocyles n_saturated_rings
0 0 245.041526 0.400000 2 1 1 3 13 2 0 ... 3.185866 0 0 0 1 0 1 0 0 0
1 1 298.123676 0.388889 2 2 4 3 21 1 0 ... 4.331775 2 0 2 1 1 2 0 0 0
2 2 198.115698 0.307692 2 2 3 2 15 0 0 ... 2.014719 1 0 1 1 1 2 0 0 0

3 rows × 23 columns

Merged both dataframes of dtree_san_df and dtree_san_df_dm based on their indices.

# merge dtree_san_df & dtree_san_df_dm
dtree_f_df = pd.merge(
    dtree_san_df[["index", "molecule_chembl_id", "pKi", "max_phase"]],
    dtree_san_df_dm,
    left_index=True,
    right_index=True
)

Checking final dataframe to make sure there were 481 rows (also that index_x and index_y were identical) and also there was an increased number of columns (columns combined from both dataframes). So this finally seemed to work.

print(dtree_f_df.shape)
dtree_f_df.head(3)
(481, 27)
index_x molecule_chembl_id pKi max_phase index_y mw fsp3 n_lipinski_hba n_lipinski_hbd n_rings ... sas n_aliphatic_carbocycles n_aliphatic_heterocyles n_aliphatic_rings n_aromatic_carbocycles n_aromatic_heterocyles n_aromatic_rings n_saturated_carbocycles n_saturated_heterocyles n_saturated_rings
0 0 CHEMBL60745 8.787812 NaN 0 245.041526 0.400000 2 1 1 ... 3.185866 0 0 0 1 0 1 0 0 0
1 1 CHEMBL208599 10.585027 NaN 1 298.123676 0.388889 2 2 4 ... 4.331775 2 0 2 1 1 2 0 0 0
2 2 CHEMBL95 6.821023 4.0 2 198.115698 0.307692 2 2 3 ... 2.014719 1 0 1 1 1 2 0 0 0

3 rows × 27 columns

The two index columns (“index_x” and “index_y”) were removed, which brought out the final preprocessed dataframe.

# Remove index_x & index_y
dtree_f_df.drop(["index_x", "index_y"], axis = 1, inplace = True)
dtree_f_df.head(3)
molecule_chembl_id pKi max_phase mw fsp3 n_lipinski_hba n_lipinski_hbd n_rings n_hetero_atoms n_heavy_atoms ... sas n_aliphatic_carbocycles n_aliphatic_heterocyles n_aliphatic_rings n_aromatic_carbocycles n_aromatic_heterocyles n_aromatic_rings n_saturated_carbocycles n_saturated_heterocyles n_saturated_rings
0 CHEMBL60745 8.787812 NaN 245.041526 0.400000 2 1 1 3 13 ... 3.185866 0 0 0 1 0 1 0 0 0
1 CHEMBL208599 10.585027 NaN 298.123676 0.388889 2 2 4 3 21 ... 4.331775 2 0 2 1 1 2 0 0 0
2 CHEMBL95 6.821023 4.0 198.115698 0.307692 2 2 3 2 15 ... 2.014719 1 0 1 1 1 2 0 0 0

3 rows × 25 columns

I then saved this preprocessed dataframe as another file in my working directory, so that it could be used for estimating experimental errors and model building in the next post.

dtree_f_df.to_csv("ache_2d_chembl.csv")


Data preprocessing reflections

In general, the order of steps could be swapped in a more logical way. The subsections presented in this post bascially reflected my thought processes, as there were some back-and-forths. The whole data preprocessing step was probably still not thorough enough, but I’ve tried to cover as much as I could (hopefully I didn’t go overboard with it…). Also, it might still not be ideal to use Ki values this freely as mentioned in post 1 (noises in data issues).

It was mentioned in scikit-learn that for decision tree models, because of its non-parametric nature, there were not a lot of data cleaning required. However, I think that might be domain-specific, since for the purpose of drug discovery, if this step wasn’t done properly, whatever result that came out of the ML model most likely would not work and also would not reflect real-life scenarios. I was also planning on extending this series to add more trees to the model, that is, from one tree (decision tree), to multiple trees (random forests), and then hopefully move on to boosted trees (XGBoost and LightGBM). Therefore, I’d better do this data cleaning step well first to save some time later (if using the same set of data).

Next post will be about model building using scikit-learn and also a small part on estimating experimental errors on the dataset - this is going to be in post 3.

Back to top

References

Dean, R. B., and W. J. Dixon. 1951. “Simplified Statistics for Small Numbers of Observations.” Analytical Chemistry 23 (4): 636–38. https://doi.org/10.1021/ac60052a025.
Tilborg, Derek van, Alisa Alenicheva, and Francesca Grisoni. 2022. “Exposing the Limitations of Molecular Machine Learning with Activity Cliffs.” Journal of Chemical Information and Modeling 62 (23): 5938–51. https://doi.org/10.1021/acs.jcim.2c01073.
]]> Machine learning projects Tree models Data preprocessing ChEMBL database Pandas RDKit Datamol Python https://jhylin.github.io/Data_in_life_blog/posts/16_ML2-1_Decision_tree/2_data_prep_tran.html Mon, 18 Sep 2023 12:00:00 GMT Decision tree Jennifer HY Lin https://jhylin.github.io/Data_in_life_blog/posts/16_ML2-1_Decision_tree/3_model_build.html Post updated on 28th April 2024 - dtreeviz code (under the “Model building” section) were updated to improve the scale of the dtreeviz tree plot


Data source

The data used here was extracted from ChEMBL database by using ChEMBL web resource client in Python. The details of all the steps taken to reach the final .csv file could be seen in these earlier posts - post 1 and post 2 (yes, it took quite a while to clean the data, so it was splitted into two posts).

  • The final .csv file used to train the model was named, “ache_2d_chembl.csv”

  • The earlier version without any RDKit 2D descriptors calculated was named, “ache_chembl.csv”

  • Both files should be in a GitHub repository called, “ML2-1_decision_tree” or in my blog repository, under “posts” folder (look for “16_ML2-1_Decision_tree” folder)


Estimate experimental errors

This part was about estimating the impact of experimental errors (pKi values) on the predictive machine learning (ML) models. It was also needed to estimate the maximum possible correlation that could be drawn from the dataset prepared from the previous two posts. I supposed it made more sense if this was done prior to building the ML model, so this wouldn’t be forgotten or missed, as we know that real-life is full of many imperfections.

This subsection was inspired by Pat Walters’ posts, which have discussed about estimating errors for experimental data with code links available in these posts:

To get started, all the required libraries were loaded as below.

import pandas as pd
import numpy as np
from sklearn.metrics import r2_score
import seaborn as sns
from sklearn import tree

Imported the preprocessed data from previous posts.

# Import data
dtree = pd.read_csv("ache_2d_chembl.csv")
dtree.drop(columns = ["Unnamed: 0"], inplace=True)
dtree.head()
molecule_chembl_id pKi max_phase mw fsp3 n_lipinski_hba n_lipinski_hbd n_rings n_hetero_atoms n_heavy_atoms ... sas n_aliphatic_carbocycles n_aliphatic_heterocyles n_aliphatic_rings n_aromatic_carbocycles n_aromatic_heterocyles n_aromatic_rings n_saturated_carbocycles n_saturated_heterocyles n_saturated_rings
0 CHEMBL60745 8.787812 NaN 245.041526 0.400000 2 1 1 3 13 ... 3.185866 0 0 0 1 0 1 0 0 0
1 CHEMBL208599 10.585027 NaN 298.123676 0.388889 2 2 4 3 21 ... 4.331775 2 0 2 1 1 2 0 0 0
2 CHEMBL95 6.821023 4.0 198.115698 0.307692 2 2 3 2 15 ... 2.014719 1 0 1 1 1 2 0 0 0
3 CHEMBL173309 7.913640 NaN 694.539707 0.666667 8 0 2 8 50 ... 2.803680 0 0 0 2 0 2 0 0 0
4 CHEMBL1128 6.698970 4.0 201.092042 0.400000 2 1 1 3 13 ... 3.185866 0 0 0 1 0 1 0 0 0

5 rows × 25 columns

The pKi column was used in the code below as it contained the experimental values (calculated from measured Ki values, usually derived from countless lab experiments) collected from different scientific literatures or other sources as stated in ChEMBL. The aim was to simulate pKi values with experimental errors added to them.

Code used for the rest of the subsection were adapted with thanks from Pat Walters’ “maximum_correlation.ipynb” with my own added comments for further explanations

# Save exp data (pKi) as an object
data = dtree["pKi"]
# Save the object as a list
data_ls = [data]

# Trial 3-, 5- & 10-fold errors
for fold in (3, 5, 10):
    # Retrieve error samples randomly from a normal distribution
    # Bewteen 0 and log10 of number-fold 
    # for the length of provided data only
    error = np.random.normal(0, np.log10(fold), len(data))
    data_ls.append(error + data)

# Convert data_ls to dataframe
dtree_err = pd.DataFrame(data_ls)
# Re-align dataframe (switch column header & index)
dtree_err = dtree_err.transpose()
# Rename columns
dtree_err.columns = ["pKi", "3-fold", "5-fold", "10-fold"]
print(dtree_err.shape)
dtree_err.head()
(481, 4)
pKi 3-fold 5-fold 10-fold
0 8.787812 8.710912 9.101193 7.471251
1 10.585027 10.883334 10.291557 9.455301
2 6.821023 6.134753 6.799967 7.122006
3 7.913640 8.390146 7.874722 7.209130
4 6.698970 7.359148 7.290723 5.770489

Melting the created dtree_err so it could be plotted later (noticed there should be an increased number of rows after re-stacking the data).

# Melt the dtree_err dataframe 
# to make error values in one column (for plotting)
dtree_err_melt = dtree_err.melt(id_vars = "pKi")
print(dtree_err_melt.shape)
dtree_err_melt.head()
(1443, 3)
pKi variable value
0 8.787812 3-fold 8.710912
1 10.585027 3-fold 10.883334
2 6.821023 3-fold 6.134753
3 7.913640 3-fold 8.390146
4 6.698970 3-fold 7.359148

Presenting this in regression plots.

Note: There was a matplotlib bug which would always show a tight_layout user warning for FacetGrid plots in seaborn (the lmplot used below). Seaborn was built based on matplotlib so unsurprisingly this occurred (this GitHub issue link might explain it). I have therefore temporarily silenced this user warning for the sake of post publication.

# To silence the tight-layout user warning
import warnings
warnings.filterwarnings("ignore")

# variable = error-fold e.g. 3-fold
# value = pKi value plus error
sns.set_theme(font_scale = 1.5)
plot = sns.lmplot(
    x = "pKi", 
    y = "value", 
    col = "variable", 
    data = dtree_err_melt, 
    # alpha = mark’s opacity (low - more transparent)
    # s = mark size (increase with higher number)
    scatter_kws = dict(alpha = 0.5, s = 15)
    )
title_list = ["3-fold", "5-fold", "10-fold"]
for i in range(0, 3):
    plot.axes[0, i].set_ylabel("pKi + error")
    plot.axes[0, i].set_title(title_list[i])

Simulating the impact of error on the correlation between experimental pKi and also pKi with errors (3-fold, 5-fold and 10-fold). R2 calculated using scikit-learn was introduced in the code below.

# Calculating r2 score (coefficient of determination) 
# based on 1000 trials for each fold
# note: data = dtree["pKi"]

# Create an empty list for correlation
cor_ls = []
for fold in [3, 5, 10]:
    # Set up 1000 trials
    for i in range(0, 1000):
        error = np.random.normal(0, np.log10(fold), len(data))
        cor_ls.append([r2_score(data, data + error), f"{fold}-fold"])

# Convert cor_ls into dataframe
err_df = pd.DataFrame(cor_ls, columns = ["r2", "fold_error"])
err_df.head()
r2 fold_error
0 0.899830 3-fold
1 0.890129 3-fold
2 0.895605 3-fold
3 0.880796 3-fold
4 0.887374 3-fold

Plotting the R2 and fold-errors as violin plots.

sns.set_theme(rc = {"figure.figsize": (9, 8)}, font_scale = 1.5)
vplot = sns.violinplot(x = "fold_error", y = "r2", data = err_df)
vplot.set(xlabel = "Fold error", ylabel = "R$^2$")
[Text(0.5, 0, 'Fold error'), Text(0, 0.5, 'R$^2$')]

This definitely helped a lot with visualising the estimated errors for the experimental Ki values curated in ChEMBL for this specific protein target (CHEMBL220, acetylcholinesterase (AChE)). The larger the error-fold, the lower the R2, and once the experimental error reached 10-fold, we could see an estimated R2 (maximum correlation) with its median sitting below 0.55, indicating a likely poor predictive ML model if it was built based on these data with the estimated 10-fold experimental errors.


Check max phase distribution

At this stage, I’ve planned to do model training on compounds with max phase 4 (i.e. prescription medicines), so this would somewhat be an attempt to mirror real-life scenarios for the ML prediction model.

Max phases were assigned to each ChEMBL-curated compound according to this ChEMBL FAQ link (under the question of “What is max phase?”). As quoted from this ChEMBL FAQ link, a max phase 4 compound means:

“Approved (4): A marketed drug e.g. AMINOPHYLLINE (CHEMBL1370561) is an FDA approved drug for treatment of asthma.”

Checking out the actual counts of each max phase group in the dataset.

dtree[["molecule_chembl_id", "max_phase"]].groupby("max_phase").count()
molecule_chembl_id
max_phase
-1.0 2
2.0 2
3.0 1
4.0 10

There was only a very small number of compounds with max phase 4 assigned (a total count of 10, which was also unsurprising since there weren’t many AChE inhibitors used as prescription medications for dementia - some of the well-known examples were donepezil, galantamine and rivastigmine).

Filling in actual “null” labels for all “NaN” rows in the “max_phase” columns to help with filtering out these compounds later on.

dtree["max_phase"].fillna("null", inplace=True)
dtree.head()
molecule_chembl_id pKi max_phase mw fsp3 n_lipinski_hba n_lipinski_hbd n_rings n_hetero_atoms n_heavy_atoms ... sas n_aliphatic_carbocycles n_aliphatic_heterocyles n_aliphatic_rings n_aromatic_carbocycles n_aromatic_heterocyles n_aromatic_rings n_saturated_carbocycles n_saturated_heterocyles n_saturated_rings
0 CHEMBL60745 8.787812 null 245.041526 0.400000 2 1 1 3 13 ... 3.185866 0 0 0 1 0 1 0 0 0
1 CHEMBL208599 10.585027 null 298.123676 0.388889 2 2 4 3 21 ... 4.331775 2 0 2 1 1 2 0 0 0
2 CHEMBL95 6.821023 4.0 198.115698 0.307692 2 2 3 2 15 ... 2.014719 1 0 1 1 1 2 0 0 0
3 CHEMBL173309 7.913640 null 694.539707 0.666667 8 0 2 8 50 ... 2.803680 0 0 0 2 0 2 0 0 0
4 CHEMBL1128 6.698970 4.0 201.092042 0.400000 2 1 1 3 13 ... 3.185866 0 0 0 1 0 1 0 0 0

5 rows × 25 columns


Sanity check on the dataframe

This was just another sanity check for myself on the dtree dataframe - making sure there weren’t any “NaN” cells in it (so dropping any “NaN” again, even though I might have already done this as one of the steps during data preprocessing).

dtree.dropna()
print(dtree.shape)
dtree.head()
(481, 25)
molecule_chembl_id pKi max_phase mw fsp3 n_lipinski_hba n_lipinski_hbd n_rings n_hetero_atoms n_heavy_atoms ... sas n_aliphatic_carbocycles n_aliphatic_heterocyles n_aliphatic_rings n_aromatic_carbocycles n_aromatic_heterocyles n_aromatic_rings n_saturated_carbocycles n_saturated_heterocyles n_saturated_rings
0 CHEMBL60745 8.787812 null 245.041526 0.400000 2 1 1 3 13 ... 3.185866 0 0 0 1 0 1 0 0 0
1 CHEMBL208599 10.585027 null 298.123676 0.388889 2 2 4 3 21 ... 4.331775 2 0 2 1 1 2 0 0 0
2 CHEMBL95 6.821023 4.0 198.115698 0.307692 2 2 3 2 15 ... 2.014719 1 0 1 1 1 2 0 0 0
3 CHEMBL173309 7.913640 null 694.539707 0.666667 8 0 2 8 50 ... 2.803680 0 0 0 2 0 2 0 0 0
4 CHEMBL1128 6.698970 4.0 201.092042 0.400000 2 1 1 3 13 ... 3.185866 0 0 0 1 0 1 0 0 0

5 rows × 25 columns


Model building
Training data based on max phase 4 compounds

So here I wanted to separate the collected data by splitting the compounds into two groups based on their assigned max phases. Compounds with max phase 4 were chosen as the training data, and the rest of the compounds with max phases of “null” would be the testing data.

# Create a df for compounds with max phase 4 only
dtree_mp4 = dtree[dtree["max_phase"] == 4]
dtree_mp4
molecule_chembl_id pKi max_phase mw fsp3 n_lipinski_hba n_lipinski_hbd n_rings n_hetero_atoms n_heavy_atoms ... sas n_aliphatic_carbocycles n_aliphatic_heterocyles n_aliphatic_rings n_aromatic_carbocycles n_aromatic_heterocyles n_aromatic_rings n_saturated_carbocycles n_saturated_heterocyles n_saturated_rings
2 CHEMBL95 6.821023 4.0 198.115698 0.307692 2 2 3 2 15 ... 2.014719 1 0 1 1 1 2 0 0 0
4 CHEMBL1128 6.698970 4.0 201.092042 0.400000 2 1 1 3 13 ... 3.185866 0 0 0 1 0 1 0 0 0
6 CHEMBL640 6.000000 4.0 235.168462 0.461538 4 3 1 4 17 ... 1.791687 0 0 0 1 0 1 0 0 0
9 CHEMBL502 7.688246 4.0 379.214744 0.458333 4 0 4 4 28 ... 2.677222 1 1 2 2 0 2 0 1 1
131 CHEMBL481 7.296709 4.0 586.279135 0.515152 10 1 7 10 43 ... 3.632560 0 4 4 1 2 3 0 2 2
133 CHEMBL360055 4.431798 4.0 510.461822 0.800000 6 0 1 6 36 ... 3.257653 0 0 0 1 0 1 0 0 0
160 CHEMBL1025 5.221849 4.0 184.066459 1.000000 3 0 0 5 11 ... 3.345144 0 0 0 0 0 0 0 0 0
171 CHEMBL659 6.522879 4.0 287.152144 0.529412 4 1 4 4 21 ... 4.226843 1 2 3 1 0 1 0 0 0
180 CHEMBL1200970 4.607303 4.0 348.142697 0.368421 2 0 3 4 23 ... 4.223591 0 1 1 2 0 2 0 0 0
195 CHEMBL1677 6.995679 4.0 234.092376 0.307692 2 2 3 3 16 ... 3.218715 1 0 1 1 1 2 0 0 0

10 rows × 25 columns

Making sure donepezil and galantamine were in this dtree_mp4 dataframe, so the model training would be based on these medicines and also other max phase 4 AChE inhibitors.

The screenshots of both medicines were taken from ChEMBL website:

Screenshot of donepezil (parent drug form) with its molecule ChEMBL ID

Screenshot of galantamine (parent drug form) with its molecule ChEMBL ID

The following regex string check confirmed that these two compounds were in the dtree_mp4 dataframe - row indices 9 and 171 contained these two drugs.

list_ache_inh = ["CHEMBL502", "CHEMBL659"]
dtree_mp4["molecule_chembl_id"].str.contains(r"|".join(list_ache_inh))
2      False
4      False
6      False
9       True
131    False
133    False
160    False
171     True
180    False
195    False
Name: molecule_chembl_id, dtype: bool

Setting up the features for the training set.

# Set X (features) for max phase 4 compounds
X_mp4_df = dtree_mp4[['mw', 'fsp3', 'n_lipinski_hba', 'n_lipinski_hbd', 'n_rings', 'n_hetero_atoms', 'n_heavy_atoms', 'n_rotatable_bonds', 'n_radical_electrons', 'tpsa', 'qed', 'clogp', 'sas', 'n_aliphatic_carbocycles', 'n_aliphatic_heterocyles', 'n_aliphatic_rings', 'n_aromatic_carbocycles', 'n_aromatic_heterocyles', 'n_aromatic_rings', 'n_saturated_carbocycles', 'n_saturated_heterocyles', 'n_saturated_rings']]

# Convert X_mp4_df to numpy array
X_mp4 = X_mp4_df.to_numpy()
X_mp4
array([[ 1.98115698e+02,  3.07692308e-01,  2.00000000e+00,
         2.00000000e+00,  3.00000000e+00,  2.00000000e+00,
         1.50000000e+01,  0.00000000e+00,  0.00000000e+00,
         3.89100000e+01,  7.06488238e-01,  2.69580000e+00,
         2.01471913e+00,  1.00000000e+00,  0.00000000e+00,
         1.00000000e+00,  1.00000000e+00,  1.00000000e+00,
         2.00000000e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00],
       [ 2.01092042e+02,  4.00000000e-01,  2.00000000e+00,
         1.00000000e+00,  1.00000000e+00,  3.00000000e+00,
         1.30000000e+01,  2.00000000e+00,  0.00000000e+00,
         2.02300000e+01,  6.08112327e-01, -1.01700000e+00,
         3.18586632e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00,  1.00000000e+00,  0.00000000e+00,
         1.00000000e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00],
       [ 2.35168462e+02,  4.61538462e-01,  4.00000000e+00,
         3.00000000e+00,  1.00000000e+00,  4.00000000e+00,
         1.70000000e+01,  6.00000000e+00,  0.00000000e+00,
         5.83600000e+01,  7.31539693e-01,  1.34040000e+00,
         1.79168720e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00,  1.00000000e+00,  0.00000000e+00,
         1.00000000e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00],
       [ 3.79214744e+02,  4.58333333e-01,  4.00000000e+00,
         0.00000000e+00,  4.00000000e+00,  4.00000000e+00,
         2.80000000e+01,  6.00000000e+00,  0.00000000e+00,
         3.87700000e+01,  7.47461492e-01,  4.36110000e+00,
         2.67722173e+00,  1.00000000e+00,  1.00000000e+00,
         2.00000000e+00,  2.00000000e+00,  0.00000000e+00,
         2.00000000e+00,  0.00000000e+00,  1.00000000e+00,
         1.00000000e+00],
       [ 5.86279135e+02,  5.15151515e-01,  1.00000000e+01,
         1.00000000e+00,  7.00000000e+00,  1.00000000e+01,
         4.30000000e+01,  4.00000000e+00,  0.00000000e+00,
         1.14200000e+02,  3.55955569e-01,  4.09110000e+00,
         3.63256044e+00,  0.00000000e+00,  4.00000000e+00,
         4.00000000e+00,  1.00000000e+00,  2.00000000e+00,
         3.00000000e+00,  0.00000000e+00,  2.00000000e+00,
         2.00000000e+00],
       [ 5.10461822e+02,  8.00000000e-01,  6.00000000e+00,
         0.00000000e+00,  1.00000000e+00,  6.00000000e+00,
         3.60000000e+01,  2.10000000e+01,  0.00000000e+00,
         2.76900000e+01,  2.05822189e-01,  5.45250000e+00,
         3.25765349e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00,  1.00000000e+00,  0.00000000e+00,
         1.00000000e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00],
       [ 1.84066459e+02,  1.00000000e+00,  3.00000000e+00,
         0.00000000e+00,  0.00000000e+00,  5.00000000e+00,
         1.10000000e+01,  4.00000000e+00,  0.00000000e+00,
         3.55300000e+01,  6.29869319e-01,  2.91400000e+00,
         3.34514393e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00],
       [ 2.87152144e+02,  5.29411765e-01,  4.00000000e+00,
         1.00000000e+00,  4.00000000e+00,  4.00000000e+00,
         2.10000000e+01,  1.00000000e+00,  0.00000000e+00,
         4.19300000e+01,  8.00524269e-01,  1.85030000e+00,
         4.22684283e+00,  1.00000000e+00,  2.00000000e+00,
         3.00000000e+00,  1.00000000e+00,  0.00000000e+00,
         1.00000000e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00],
       [ 3.48142697e+02,  3.68421053e-01,  2.00000000e+00,
         0.00000000e+00,  3.00000000e+00,  4.00000000e+00,
         2.30000000e+01,  5.00000000e+00,  0.00000000e+00,
         6.48000000e+00,  7.09785317e-01,  5.44140000e+00,
         4.22359068e+00,  0.00000000e+00,  1.00000000e+00,
         1.00000000e+00,  2.00000000e+00,  0.00000000e+00,
         2.00000000e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00],
       [ 2.34092376e+02,  3.07692308e-01,  2.00000000e+00,
         2.00000000e+00,  3.00000000e+00,  3.00000000e+00,
         1.60000000e+01,  0.00000000e+00,  0.00000000e+00,
         3.89100000e+01,  7.60853221e-01,  3.11760000e+00,
         3.21871482e+00,  1.00000000e+00,  0.00000000e+00,
         1.00000000e+00,  1.00000000e+00,  1.00000000e+00,
         2.00000000e+00,  0.00000000e+00,  0.00000000e+00,
         0.00000000e+00]])

Setting up the target for the training set.

# Set y (target) for max phase 4 compounds
y_mp4_df = dtree_mp4["pKi"]

# Convert y_mp4_df to numpy array
y_mp4 = y_mp4_df.to_numpy()
y_mp4
array([6.82102305, 6.69897   , 6.        , 7.68824614, 7.29670862,
       4.43179828, 5.22184875, 6.52287875, 4.60730305, 6.99567863])

The DecisionTreeRegressor() was fitted on the compounds with max phase 4 as shown below, keeping tree depth at 3 for now to avoid complicating the overall tree graph (the deeper the tree, the more branches - potentially might overfit and create noises for the model).

ache_tree_mp4 = tree.DecisionTreeRegressor(max_depth=3, random_state=1)
ache_tree_mp4 = ache_tree_mp4.fit(X_mp4, y_mp4)


scikit-learn tree plot

A simple decision tree plot based on scikit-learn’s plot_tree() was shown below.

tree.plot_tree(ache_tree_mp4, feature_names=list(X_mp4_df.columns), filled=True, rounded=True)
[Text(0.5909090909090909, 0.875, 'clogp <= 4.901\nsquared_error = 1.144\nsamples = 10\nvalue = 6.228'),
 Text(0.36363636363636365, 0.625, 'n_aromatic_rings <= 1.5\nsquared_error = 0.516\nsamples = 8\nvalue = 6.656'),
 Text(0.18181818181818182, 0.375, 'n_aromatic_rings <= 0.5\nsquared_error = 0.33\nsamples = 4\nvalue = 6.111'),
 Text(0.09090909090909091, 0.125, 'squared_error = 0.0\nsamples = 1\nvalue = 5.222'),
 Text(0.2727272727272727, 0.125, 'squared_error = 0.088\nsamples = 3\nvalue = 6.407'),
 Text(0.5454545454545454, 0.375, 'n_rotatable_bonds <= 2.0\nsquared_error = 0.108\nsamples = 4\nvalue = 7.2'),
 Text(0.45454545454545453, 0.125, 'squared_error = 0.008\nsamples = 2\nvalue = 6.908'),
 Text(0.6363636363636364, 0.125, 'squared_error = 0.038\nsamples = 2\nvalue = 7.492'),
 Text(0.8181818181818182, 0.625, 'mw <= 429.302\nsquared_error = 0.008\nsamples = 2\nvalue = 4.52'),
 Text(0.7272727272727273, 0.375, 'squared_error = 0.0\nsamples = 1\nvalue = 4.607'),
 Text(0.9090909090909091, 0.375, 'squared_error = -0.0\nsamples = 1\nvalue = 4.432')]

scikit-learn tree plot in texts

It could also be in the text form.

from sklearn.tree import export_text

text = export_text(ache_tree_mp4, feature_names=list(X_mp4_df.columns))
print(text)
|--- clogp <= 4.90
|   |--- n_aromatic_rings <= 1.50
|   |   |--- n_aromatic_rings <= 0.50
|   |   |   |--- value: [5.22]
|   |   |--- n_aromatic_rings >  0.50
|   |   |   |--- value: [6.41]
|   |--- n_aromatic_rings >  1.50
|   |   |--- n_rotatable_bonds <= 2.00
|   |   |   |--- value: [6.91]
|   |   |--- n_rotatable_bonds >  2.00
|   |   |   |--- value: [7.49]
|--- clogp >  4.90
|   |--- mw <= 429.30
|   |   |--- value: [4.61]
|   |--- mw >  429.30
|   |   |--- value: [4.43]


graphviz tree plot

The graphviz version, which showed a small variation in graph presentation, seemed to be much larger in size and easier to view.

import graphviz

dot_data = tree.export_graphviz(ache_tree_mp4, out_file=None, feature_names=list(X_mp4_df.columns), filled=True, rounded=True, special_characters=False)

graph = graphviz.Source(dot_data)
graph


dtreeviz tree plot

The following was a dtreeviz version of the decision tree, which actually included the regression plots of different molecular features e.g. clogp versus the target value of pKi. It seemed a bit more intuitive as these plots clearly showed where the threshold cut-offs would be for each feature (molecular descriptors). The GitHub repository link for dtreeviz could be accessed here.

import dtreeviz 

viz = dtreeviz.model(ache_tree_mp4, X_train=X_mp4, y_train=y_mp4, target_name="pKi", feature_names=list(X_mp4_df.columns))
# Added "scale = 2" to view()
# to make plot larger in size
viz.view(scale = 2)


Testing and predicting data based on max phase of null compounds
# Compounds with max phase as "null"
dtree_mp_null = dtree[dtree["max_phase"] == "null"]
print(dtree_mp_null.shape)
dtree_mp_null.head() 
(466, 25)
molecule_chembl_id pKi max_phase mw fsp3 n_lipinski_hba n_lipinski_hbd n_rings n_hetero_atoms n_heavy_atoms ... sas n_aliphatic_carbocycles n_aliphatic_heterocyles n_aliphatic_rings n_aromatic_carbocycles n_aromatic_heterocyles n_aromatic_rings n_saturated_carbocycles n_saturated_heterocyles n_saturated_rings
0 CHEMBL60745 8.787812 null 245.041526 0.400000 2 1 1 3 13 ... 3.185866 0 0 0 1 0 1 0 0 0
1 CHEMBL208599 10.585027 null 298.123676 0.388889 2 2 4 3 21 ... 4.331775 2 0 2 1 1 2 0 0 0
3 CHEMBL173309 7.913640 null 694.539707 0.666667 8 0 2 8 50 ... 2.803680 0 0 0 2 0 2 0 0 0
5 CHEMBL102226 4.698970 null 297.152928 0.923077 3 0 0 5 18 ... 2.965170 0 0 0 0 0 0 0 0 0
7 CHEMBL103873 5.698970 null 269.121628 0.909091 3 0 0 5 16 ... 3.097106 0 0 0 0 0 0 0 0 0

5 rows × 25 columns

There were 466 compounds with max phase as “null”, meaning they were pre-clinical compounds. This was confirmed through the answer from ChEMBL FAQ link, a max phase of “null” compound means:

“Preclinical (NULL): preclinical compounds with bioactivity data e.g. is a preclinical compound with bioactivity data that has been extracted from scientific literature. However, the sources of drug and clinical candidate drug information in ChEMBL do not show that this compound has reached clinical trials and therefore the max_phase is set to null.”

Again, setting up the features for the testing dataset.

X_mp_test_df = dtree_mp_null[['mw', 'fsp3', 'n_lipinski_hba', 'n_lipinski_hbd', 'n_rings', 'n_hetero_atoms', 'n_heavy_atoms', 'n_rotatable_bonds', 'n_radical_electrons', 'tpsa', 'qed', 'clogp', 'sas', 'n_aliphatic_carbocycles', 'n_aliphatic_heterocyles', 'n_aliphatic_rings', 'n_aromatic_carbocycles', 'n_aromatic_heterocyles', 'n_aromatic_rings', 'n_saturated_carbocycles', 'n_saturated_heterocyles', 'n_saturated_rings']]

# Convert X_mp_test_df to numpy array
X_mp_test = X_mp_test_df.to_numpy()
X_mp_test
array([[2.45041526e+02, 4.00000000e-01, 2.00000000e+00, ...,
        0.00000000e+00, 0.00000000e+00, 0.00000000e+00],
       [2.98123676e+02, 3.88888889e-01, 2.00000000e+00, ...,
        0.00000000e+00, 0.00000000e+00, 0.00000000e+00],
       [6.94539707e+02, 6.66666667e-01, 8.00000000e+00, ...,
        0.00000000e+00, 0.00000000e+00, 0.00000000e+00],
       ...,
       [3.11152144e+02, 3.15789474e-01, 4.00000000e+00, ...,
        0.00000000e+00, 0.00000000e+00, 0.00000000e+00],
       [3.68096076e+02, 9.23076923e-01, 4.00000000e+00, ...,
        0.00000000e+00, 2.00000000e+00, 2.00000000e+00],
       [2.46136828e+02, 5.00000000e-01, 4.00000000e+00, ...,
        0.00000000e+00, 3.00000000e+00, 3.00000000e+00]])

Then setting up the target for the testing set.

y_test = ache_tree_mp4.predict(X_mp_test)

The trained model, ache_tree_mp4, was used to predict on the testing dataset (max phase of null compounds).

ache_tree_mp4 = ache_tree_mp4.fit(X_mp_test, y_test)

Using the graphviz graph version to show the decision tree on the testing set, as the prediction result.

dot_data = tree.export_graphviz(ache_tree_mp4, out_file=None, feature_names=list(X_mp_test_df.columns), filled=True, rounded=True, special_characters=False)

graph = graphviz.Source(dot_data)
graph


Discussions

Because of the small amount of training data, this might hint at using an ensemble approach in the future, where model averaging would be derived from a bunch of tree models rather than using a single tree model, which was what I did here. The reason I started with one tree was because it was no point in building a multiple-tree model if one had no clue about how one tree was built. To learn as much as possible, I had to dissect the multiple-tree version first to focus on one tree at a time.

One thing I’d like to mention was that rivastigmine was not included in the training dataset because it was actually not a pure AChE inhibitor (as it was also a butyrylcholinesterase (BChE) inhibitor), since my focus was purely on AChE at this time, this particular drug was unfortunately excluded. However, I did make sure the other two drugs (donepezil and galantamine) were included in the training dataset. One possible thing to do in the future if I want to improve this was to add BChE as another protein target and perhaps add this as an additional dataset towards the model.

As described in the subsection of “Estimate experimental errors”, there were experimental errors of 3-fold, 5-fold and 10-fold estimated based on the provided pKi data. With the prediction model used in this post, the estimated experimental errors would need to be taken into consideration, particularly at the time when the model was being investigated during the model evaluation and validation step (however due to the length of series 2.1 posts, I decided not to add this step yet, but would try to look at this later in the multiple tree model series if this applies).

A bit about the last decision tree plot, tentatively clogp (calculated partition coefficient) might be the crucial molecular feature in deciding whether a molecule might be closer to being an AChE inhibitor. Other important molecular features also included the number of aromatic rings, molecular weights, solvent accessible surface area and others (I’ve set the random state to 1 for now, so hopefully the result will be reproducible as I realised my old code without it always generated different tree plots, then all of the sudden I remembered that I forgot to set the random state of the estimator, so this was added).

Since the type of AChE inhibitors was not the focus of this series, I won’t go into details about which value of pKi or Ki would lead to the ideal AChE inhibitor (the well-known Cheng-Prusoff equation (Cheng and Prusoff 1973) might also lead to many discussions about Ki and IC50 values). This is because there are at least two types of AChE inhibitors available - reversible and irreversible (Colovic et al. 2013). Donepezil, galantamine and rivastigmine mentioned previously are the commonly known reversible AChE inhibitors. The irreversible type, as the name suggested, is usually used as insecticides or nerve agents. Another reason is that I didn’t go into details checking all of the identities for the 10 max phase 4 compounds used in the training set, as I only really made sure that donepezil and galantamine were included in the 10 molecules. If I were to re-model again purely on reversible AChE inhibitors targeting dementia or Alzheimer’s disease, I think I had to be quite sure of what I was training the model with, i.e. excluding irreversible AChE inhibitors from the training set.

However, if our aim was to only find novel AChE inhibitors in a general manner, one of the ways to check post-model building would be to re-run the dataframe again on compounds with max phase as null, including the molecular feature names to find out which compounds were at the predicted threshold cut-off values to see if their corresponding pKi values (note: these preclinical compounds had Ki values extracted from literature sources etc.) would match the predicted ones. One caveat of this method was that there might be pre-existing experimental errors in all the obtained and recorded Ki values, so this might not confirm that the model was truly a good reflection of the real-life scenario. Therefore, at most, this would probably add a very small value during the model evaluation phase.

The best way would be to test all of these compounds in the same experimental set-ups, through same experimental steps, and in the same laboratory to find out their respective Ki (or pKi) values. However, this was most likely not very feasible due to various real-life restrictions (the availability of financial and staffing resources). The most likely outcome might be to choose a selected group of compound candidates with the highest possibilities to proceed in the drug discovery pipeline based on past experimental, ML and clinical experiences, and then test them in the ways mentioned here.

I also came across a blog post about calculating the prediction intervals of ML prediction models (which mentioned the MAPIE package), but I didn’t quite get time to look into this package yet, and from what I have read in its repository link, it potentially could be quite useful for classification, regression and time-series models.


Final words

I didn’t think a definite conclusion could be drawn here, as this was only purely from one very simple and single decision tree, so I have named this last part as “final words”, as I felt if I didn’t stop here, this post or series of posts could go on forever or as long it could. The main thing here was to fully understand how one single decision tree was constructed based on hopefully reasonable-ish data (still not the best as I could not rule out all the noises from the data), and then to view the tree visually in different styles of plots. It was also important to understand how this was a white-box ML approach with clear features or descriptions shown to trace where the tree would branch off to reach different final outcomes or targets. This series was really a preamble for the multiple-tree models e.g. random forest and boosted trees, as I have bravely planned to do a series of posts on tree models due to my interests in them, so that might take a while, slowly but hopefully surely.


Acknowledgements

I’d like to thank all the authors for all the open-source packages used in the series 2.1 posts. I’d also like to thank all the authors of all the blog posts mentioned in this series as well since I’ve learnt a lot from them too.

Back to top

References

Brown, Scott P., Steven W. Muchmore, and Philip J. Hajduk. 2009. “Healthy Skepticism: Assessing Realistic Model Performance.” Drug Discovery Today 14 (7-8): 420–27. https://doi.org/10.1016/j.drudis.2009.01.012.
Cheng, Yung-Chi, and William H. Prusoff. 1973. “Relationship Between the Inhibition Constant (KI) and the Concentration of Inhibitor Which Causes 50 Per Cent Inhibition (I50) of an Enzymatic Reaction.” Biochemical Pharmacology 22 (23): 3099–3108. https://doi.org/10.1016/0006-2952(73)90196-2.
Colovic, Mirjana B., Danijela Z. Krstic, Tamara D. Lazarevic-Pasti, Aleksandra M. Bondzic, and Vesna M. Vasic. 2013. “Acetylcholinesterase Inhibitors: Pharmacology and Toxicology.” Current Neuropharmacology 11 (3): 315–35. https://doi.org/10.2174/1570159x11311030006.
]]> Machine learning projects Tree models Pandas Scikit-learn ChEMBL database Python https://jhylin.github.io/Data_in_life_blog/posts/16_ML2-1_Decision_tree/3_model_build.html Mon, 18 Sep 2023 12:00:00 GMT Molecular visualisation (Molviz) web application Jennifer HY Lin https://jhylin.github.io/Data_in_life_blog/posts/15_Molviz/Molviz.html The final deployed app is on Shinyapps.io:

Update on 29/7/2026: I’ve come across the news update recently that Posit is slowly going to phase out several of its different platforms including shinyapps.io and start integrating them into one uniformed platform. I’ve gone back to visit my old HF account and realised that the app is actually running fine currently (!), except that sometimes the images of molecules may get over blown that’ll cover some of the other surrounding texts underneath. So another link to this old app has been added at the top. Perhaps I should look at updating/upgrading the app in fact if time allows…


Background - how the app started

Originally I had an idea of incorporating mols2grid library within Shiny for Python web app framework (after seeing an example of a similar app in Streamlit previously). So I worked on a few ideas, but obviously mols2grid was designed to work inside Jupyter Notebook/Lab and Shiny for Python was only out of alpha at that stage so things were still being developed. After a few trials, unfortunately mols2grid wasn’t directly compatible with the Shiny for Python framework at that time (I even wrote a small story about it as a comment to an issue).

I then went away to work on another project on molecular scaffolds and left this mini project aside. However, recently I had another idea of trying to build a Shiny for Python app from the scratch (with a focus on cheminformatics or chemical information), so that users in relevant fields can view and save 2D images of small molecules in a web browser environment instead of only inside a Jupyter Notebook/Lab. I also thought to place the Shiny for Python framework to test if it was being used in a more intensive area such as chemistry and drug discovery.

Another reason that have triggered this side project was that I came across a comment from an old RDKit blog post from someone asking about how to save compound image as a PNG1 file, since SVG2 version was hard to convert etc. (or something along that line). I thought it should be possible, and this should not be only limited to Jupyter environments only (thinking of people not doing coding at all…), so here we are.


About each version of the app

I’ll try to explain what each version of the app_x.py script entails, as there are currently several different versions of them inside the repository. The final version is the one called “app.py”, named this way so that it’ll be recognised by rsconnect/Shinyapps.io when deploying the app. By providing some explanations below should also show that it was quite a process to arrive at the final “app.py”, it wasn’t built within a day for sure (at least for me).


app_molviz_v1.py

This was the first version that purely provided the ability to show 2D images of the molecules via selecting their corresponding index numbers. The libraries used appeared less aligned and a few tests were run below (some of them commented out during trials). This was the one that I’ve figured out how to make the image appeared in the app.


app_molviz_v2.py

For the second version, I started thinking about how I would present the app in a simple layout for the end users. The backbone code to support image generations was by using rdkit.Chem.Draw package’s MolToImage() module, which normally returns a PIL3 image object, and also supports atom and bond highlighting. Another useful module that I’ve tried was MolToFile() within the same package, which would generate and save a PNG file for a specified molecule from the dataframe.

I then took a bit more time to familiarise myself with some basic PIL image manipulations, and used online resources to formulate code to merge separate PNG images into one table grid-like image - potentially may be useful for substructural or R-group comparisons.

I have also added the interactive data table at the end to see how it would fit in with the rest of the app.


app_molviz_v3.py

The third version mainly dealt with how to segregate and differentiate between highlighting or non-highlighting and also with or without index numbers showing for the compounds in the images. I’ve tried to use a different code for atom labelling this time with thanks to this link. However, there was always an issue of not being able to flip back from with index to without index, since the atom labelling code itself somehow overflows its effect to the rest after labelling the atom indices (presumably this atom labelling code would work great in a Jupyter notebook scenario).


app_molviz_v4.py & app_molviz_v5.py

Both version 4 and 5 were where I’ve tested using “atomNote” (numbers appear beside atoms) instead of “atomLabel” (numbers replaces atoms directly in structures) to label atoms in molecular structures.

An example of the atom labelling code would look like this (replace ‘atomNote’ with ‘atomLabel’ to get different labelling effect):

```{python}
for atom in mols[input.mol()].GetAtoms():
  atom.SetProp('atomNote', str(atom.GetIdx()))
```

I’ve also started adding introductory texts for the app and edited the layout a bit more.


app_molviz_v6_hf.py

This was basically the final version of the app, but with code edited to attempt to deploy the app on HuggingFace (HF). The main difference I was testing was on how to store the saved images as Docker was new to me at the time, and then while I was thinking about changing the Dockerfile, there was actually another problem in relation to the cairosvg code. Because of this, I then placed this deployment on hold in order to buy more time to figure out code, and also to try Shinyapps.io to see if this could be deployed.


app_molviz_v6.py or app.py

This was the last version and was the version used to deploy the app on Shinyapps.io. I had to rename the file as mentioned previously to “app.py” so that the Shinyapps.io servers would recognise this Python script as the script to run the app (otherwise it wouldn’t be deployed successfully, this took me a few tries and to read the log file to figure this out). So it was saved as a separate file, and for any latest text changes in the app I would refer to app.py as the most current app file.

The biggest code change was that I ended up not using the MolToImage() or MolToFile() modules, but rather I used rdMolDraw2D module from rdkit.Chem.Draw package. The reason being I’ve noticed the image resolutions weren’t great for the previously used modules (Jupyter notebook environments should not have this problem, as you could simply switch on this line of code by setting it to true like this, IPythonConsole.ipython_useSVG = True). So I resorted to other means and came across this useful link to generate images with better resolutions, and introduced the cairosvg library.

So the code was changed and would now use rdMolDraw2D.MolDraw2DSVG() first and add on addAtomIndices from drawOptions() and also DrawMolecule() to highlight substructures. The SVG generated would then be converted to PNG via cairosvg library. The end result produced slightly better image resolutions. Although I’ve found for more structurally complexed molecules, the image size would really need to be quite large to be in the high resolution zone. For compounds with simpler structures, this seemed to be much less of a problem. This was also why I had to have these PNG images blown up this large in the app, to cater for the image resolution aspect.


Other files
code_test.py

I’m not exactly sure how other data scientists/developers work, but for me since I came from a completely different background and training, I’m used to plan, set up and do experiments to test things I’d like to try, and see where the results will lead me to. So for this in a virtual computer setting, I used the “code_tests.py” to test a lot of different code.

If you go into this file, you’ll likely see a lot of RDKit code trials, and I have had a lot of fun doing this since I got to see results straight away when running the code, and learn new code and functions that way. If the end result was not the one I’ve intended, I would go on short journeys to look for answers (surprisingly I didn’t use any generative AI chatbots actually), it would be chosen intuitively as I searched online, but for this particular project, a lot of it was from past RDKit blogs, StackOverflow and random snippets I came across that have given me ideas about solving issues I came across.


app_table.py & app_itables.py

These two files were trials for incorporating dataframe inside a web app. The difference was that app_table.py was just a data table sitting inside the app, without any other particular features, while app_itables.py utilised a different package called itables, which provided an interactive feature to search for data in the table. The previous post on data source used for this app was presented as an interactive data table embedded inside a Quarto document, the same principle would also apply for a table inside a Jupyter notebook environment.


app_sample.py

This file was provided by Posit (formerly known as RStudio) from their Python for Shiny app template in HF as an example script for an app.


Features of the app

There are three main features for this app which allows viewing, saving4 and highlighting substructures of target molecules as PNG image files. I’m contemplating about adding a download feature for image file saving on the deployed app version but because I’m currently using the free version of Shinyapps.io with limited amount of data available, this may be unlikely (also because the app is more of a demonstration really as the focus is not to provide particular data/image downloads).


App deployment

There were two places I’ve tried so far, which were HF and Shinyapps.io. As mentioned briefly earlier under the subsection of “app_molviz_v6_hf.py”, it turned out cairosvg code didn’t quite play out as expected. I have so far not returned to fix this yet on HF, since I’ve managed to deploy the app on Shinyapps.io. I had a feeling I might need to revert back to the older code version with poorer image resolutions, so that was also another reason why I haven’t fixed it yet as I’d prefer to keep the better resolution one (unless someone has better ideas out there).

However, deploying the app to Shinyapps.io also wasn’t a smooth ride as well, there were some problems initially. The very first problem I got was being told by the error message that rsconnect-python was only compatible with Python version 3.7, 3.8, 3.9 and 3.10 only. I did some information digging in the Posit community forum, and I think several people mentioned using 3.9 without any problems to deploy their apps. Python version 3.11 definitely did not work at all so please avoid for now if anyone would like to try using Shiny for Python app (unless updated by rsconnect-Python in the future).

So I think the ideal app building workflow might be like this:

Note: all code examples below are to be run in the command line interface

  1. Refer to this link provided by Shiny for Python, which details about how to set up the working directory, download the Shiny package and create a virtual environment

  2. When creating the virtual environment, use venv which was already built-in within Python (and also as suggested by the Shiny for Python link) and set it to a compatible Python version.

```{python}
# To create a venv with a specific Python version e.g. Python 3.9
python3.9 -m venv my_venv_name

# Activate the created venv
source my_venv_name/bin/activate
```
  1. If you’ve accidentally set it to Python 3.11 (like what I did), just deactivate the old venv and re-create another one by using the code above. The code below can be used to deactivate the venv set up in the first place.
```{python}
# Deactivate the old venv
deactivate
```
  1. If you had to set up a new venv with a new Python version, and did not want to re-add/install all the packages or libraries used in the older version, save all the pip packages like this code below as a requirements.txt file.
```{python}
pip freeze > requirements.txt
```
  1. Once the requirements.txt was saved and after the new venv was set up and activated, install all the saved packages used to run the app by using this code.
```{python}
pip freeze -r requirements.txt
```

  1. Start coding for your app and have fun - don’t forget to save and push the files to your repository.

  2. To deploy to Shinyapps.io, follow this link, which explains all the steps. One thing I would like to remind again here is to make sure the app script (i.e. the one with data source, user interface and server code) was saved as “app.py”, so that the rsconnect-python server will recognise it and be able to deploy it to the cloud environment.


Further improvements of the app

There are of course a few things I think could be done to make the app better.

  1. It may be useful to add a download option as mentioned previously, but for demonstration purpose, I’m leaving it as a “View” only for now, unless I get comments from readers that they’d like to try this. For localhost version, the saving image function should work with files saved to the working directory.

  2. It may be even better to use SMARTS5 or SMILES for highlighting compound substructures actually (atom and bond numbering can be a bit tricky, I’ve tried the app myself, and it might not be as straight forward). I’m using atom indices here since I’m using a specific code in RDKit, but perhaps more experienced users for RDKit will know how to make code alterations etc.

  3. The app layout could be further optimised for aesthetics e.g. interactive data table could be placed at a different location, and potentially the data table could contain other data such as compound bioassay results to really fit in the structure-activity relationship exploring task.


Final words

The whole idea behind this side project was to show that interested users could use this web app framework to build an interactive app by using their own data. Other useful web app frameworks are also available out there and could potentially be more or equally useful as this one (I’m simply testing out Python for Shiny here since it’s relatively new). In a drug discovery and development setting, this could be useful to make non-coding members understand where the computer side was trying to do, and to possibly assist them during their lab workflows, hoping to add some convenience at least.


Acknowledgements

As always, I’d like to thank all of the authors of RDKit, Datamol, Shiny for Python, itables, Pandas and Polars, without them, I don’t think I can build this app out of the blue.

Back to top

Footnotes

  1. Portable network graphic (image file)↩︎

  2. Scalable vector graphic (image file)↩︎

  3. Python image library↩︎

  4. This is currently limited to localhost version if running the app.py in IDE such as VS Code, where the saved files can be located in the working directory. The deployed version on Shinyapps.io currently only allow image viewing and structure highlighting only.↩︎

  5. SMILES arbitrary target specification↩︎

]]> Python Datamol Shiny Pandas Polars itables https://jhylin.github.io/Data_in_life_blog/posts/15_Molviz/Molviz.html Wed, 09 Aug 2023 12:00:00 GMT Molecular visualisation (Molviz) web application Jennifer HY Lin https://jhylin.github.io/Data_in_life_blog/posts/15_Molviz/itables.html
Introduction

This time I’m trying to build a web application in the hope to contribute my two cents towards minimising the gap between computational and laboratory sides in a drug discovery (or chemistry-related work) setting. There are definitely many other useful web applications available out there for similar uses, but I guess each one has its own uniqueness and special features.

For this app, it is mostly aimed at lab chemists who do not use any computer programming code at all in their work, and would like to quickly view compounds in the lab while working, and also to be able to highlight compound substructures during discussions or brainstorming for ideas during compound synthesis. Importantly, this app can exist outside a Jupyter notebook environment with internet required to access the app.

This is also the first part prior to the next post which will showcase the actual app. This part mainly involves some data cleaning but not as a major focus for this post. This is not to say that data cleaning is not important, but instead they are fundamental to any work involving data in order to ensure reasonable data quality, which can then potentially influence decisions or results. I have also collapsed the code sections below to make the post easier to read (to access code used for each section, click on the “Code” links).


Code and explanations

It was actually surprisingly simple for this first part when I did it - building the interactive table. I came across this on LinkedIn on a random day for a post about itables being integrated with Shiny for Python plus Quarto. It came at the right time because I was actually trying to build this app. I quickly thought about incorporating it with the rest of the app so that users could refer back to the data quickly while visualising compound images. The code and explanations on building an interactive table for dataframes in Pandas and Polars were provided below.

To install itables, visit here for instructions and also for other information about supported notebook editors.

Code 
# Import dataframe libraries
import pandas as pd
import polars as pl

# Import Datamol
import datamol as dm

# Import itables
from itables import init_notebook_mode, show
init_notebook_mode(all_interactive=True)


# Option 1: Reading df_ai.csv as a pandas dataframe
#df = pd.read_csv("df_ai.csv")
#df.head

# Option 2: Reading df_ai.csv as a polars dataframe
df = pl.read_csv("df_ai.csv")
#df.head()


# Below was the code I used in my last post to fix the missing SMILES for neomycin 
# - the version below was edited due to recent updates in Polars
# Canonical SMILES for neomycin was extracted from PubChem 
# (https://pubchem.ncbi.nlm.nih.gov/compound/Neomycin)

df = df.with_columns([
    pl.when(pl.col("Smiles").str.lengths() == 0)
    .then(pl.lit("C1C(C(C(C(C1N)OC2C(C(C(C(O2)CN)O)O)N)OC3C(C(C(O3)CO)OC4C(C(C(C(O4)CN)O)O)N)O)O)N"))
    .otherwise(pl.col("Smiles"))
    .keep_name()
])

#df.head()

Polars dataframe library was designed without the index in mind (which is different to Pandas), therefore the itables library did not work on my specific polars dataframe which required an index column to show (note: all other Polars dataframes should work fine with itables without the index column!).

However to show row counts in Polars dataframes, we could use with_row_count() that starts the index from 0, and this would show up in a Jupyter environment as usual. A small code example would be like this below.

Code 
# Uncomment below to run
#df = df.with_row_count()

Then I converted the Polars dataframe into a Pandas one (this could be completely avoided if you started with Pandas actually).

Code 
df = df.to_pandas()

Then I added Datamol’s “_preprocess” function to convert SMILES1 into other molecular representations such as standardised SMILES (pre-processed and cleaned SMILES), SELFIES2, InChI3, InChI keys - just to provide extra information for further uses if needed. The standardised SMILES generated here would then be used for generating the molecule images later (in part 2).

Code 
# Pre-process molecules using _preprocess function - adapted from datamol.io

smiles_column = "Smiles"

dm.disable_rdkit_log()

def _preprocess(row):
    mol = dm.to_mol(row[smiles_column], ordered=True)
    mol = dm.fix_mol(mol)
    mol = dm.sanitize_mol(mol, sanifix=True, charge_neutral=False)
    mol = dm.standardize_mol(
        mol,
        disconnect_metals=False,
        normalize=True,
        reionize=True,
        uncharge=False,
        stereo=True,
    )

    row["standard_smiles"] = dm.standardize_smiles(dm.to_smiles(mol))
    row["selfies"] = dm.to_selfies(mol)
    row["inchi"] = dm.to_inchi(mol)
    row["inchikey"] = dm.to_inchikey(mol)
    return row

#Apply the _preprocess function to the prepared dataframe.
df = df.apply(_preprocess, axis = 1)
#df.head()

The next step was to keep the index column of the Pandas dataframe as an actual column (there was a reason for this, mainly for the app).

Code 
# Convert index of Pandas df into a column
df = df.reset_index()
#df.head()


Interactive data table

  • An interactive table of all the prescription-only antibiotics from ChEMBL is shown below

  • Scroll the table from left to right to see the SMILES, standardised SMILES, SELFIES, InChI, InChI keys for each compound

  • Use the search box to enter compound names to search for antibiotics and move between different pages when needed

Code 
df
index Name Smiles standard_smiles selfies inchi inchikey
Loading... (need help?) 


# Saving cleaned df_ai.csv as a new .csv file (for app_image_x.py - script to build the web app)
# df = pl.from_pandas(df)
# df.write_csv("df_ai_cleaned.csv", sep = ",")


Options for app deployment

Since I had a lot of fun deploying my previous app in Shinylive last time, I thought I might try the same this time - deploying the Molviz app as a Shinylive app in Quarto. However, it didn’t work as expected, with reason being that RDKit wasn’t written in pure Python (it was written in Python and C++), so there wasn’t a pure Python wheel file available in PyPi - this link may provide some answers relating to this. Essentially, packages or libraries used in the app will need to be compatible with Pyodide in order for the Shinylive app to work. So, the most likely option to deploy this app now would be in Shinyapps.io or HuggingFace as I read about it recently.


Next post

Code and explanations for the actual Molviz app will be detailed in the next post. To access full code and files used for now, please visit this repository link.

Back to top

Footnotes

  1. Simplified Molecular Input Line Entry Systems↩︎

  2. SELF-referencIng Embedded Strings↩︎

  3. International Chemical Identifier↩︎

]]> Python Datamol Pandas Polars itables https://jhylin.github.io/Data_in_life_blog/posts/15_Molviz/itables.html Wed, 09 Aug 2023 12:00:00 GMT Working with scaffolds in small molecules Jennifer HY Lin https://jhylin.github.io/Data_in_life_blog/posts/14_Scaffolds_in_small_molecules/chembl_anti-inf_data_prep_current.html

Features in post

This post will mainly be about the following:

  • Pre-process and standardise compounds (e.g. converting SMILES1 into SELFIES2 and other forms)

  • Obtain scaffolds for compounds

  • Align scaffolds of compounds

  • Query target scaffolds against a dataframe of compounds:

    • Function for saving multiple SMILES in .smi file

    • Function for converting .smi file into a list to query and match scaffolds of interests

    • Identify any similarities or differences in target compound of interest against other compounds in a dataframe


Quick words

I’ve always wanted to shorten my posts to a more readable length, but it was proven to be hard again, as this post was much longer than expected. Page content links are available on the right-hand side if needing to jump to sections for quick reads.


Key question to answer

Will the scaffold of compound 3 (compound of interest) be similar to the scaffolds of any approved anti-infectives in ChEMBL database?


Import libraries

The following libraries were used in this post.

import polars as pl
import pandas as pd
import datamol as dm
import mols2grid

from rdkit import Chem
from rdkit.Chem import AllChem
from rdkit.Chem.rdmolfiles import SmilesWriter, SmilesMolSupplier

# Following library was modified & adapted from  
# Patrick Walters' tutorial on "Identifying Scaffolds" 
# - links provided in "scaffold_finder library" section under
# the subheading of "Combining ChEMBL anti-infectives and FtsZ compounds"
from scaffold_finder_test import find_scaffolds, get_molecules_with_scaffold


ChEMBL anti-infectives

Data cleaning

The dataset used was extracted from ChEMBL database, with a focus on the anti-infectives.

df_ai = pl.read_csv("chembl_anti-inf.csv", sep = ";")
df_ai
shape: (144, 29)
Parent Molecule Name Synonyms Research Codes Phase Drug Applicants USAN Stem USAN Year USAN Definition USAN Stem - Substem First Approval ATC Codes Level 4 ATC Codes Level 3 ATC Codes Level 2 ATC Codes Level 1 ATC Codes Indication Class Patent Drug Type Passes Rule of Five First In Class Chirality Prodrug Oral Parenteral Topical Black Box Availability Type Smiles
str str str str f64 str str str str str str str str str str str str str str i64 i64 str i64 i64 i64 i64 i64 str str
"CHEMBL374975" "FUSIDIC ACID" "ANHYDROUS FUSI... "CEM-102|NSC-56... 4.0 "" "" "1967" "" "" "" "J01XC01 | S01A... "J01XC - Steroi... "J01X - OTHER A... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "None" "1:Synthetic Sm... 0 0 "Single Stereoi... 0 0 0 0 0 "Unknown" "CC(=O)O[C@H]1C...
"CHEMBL130" "CHLORAMPHENICO... "9-HYDROXY-9-PH... "NSC-16331|NSC-... 4.0 "Ivax Pharmaceu... "" "" "" "" "1950" "S01AA01 | S03A... "S01AA - Antibi... "S01A - ANTIINF... "S01 - OPHTHALM... "S - SENSORY OR... "Antibacterial;... "None" "1:Synthetic Sm... 1 0 "Single Stereoi... 0 1 1 1 0 "Unknown" "O=C(N[C@H](CO)...
"CHEMBL186" "CEFEPIME" "BMY-28142|CEFE... "BMY-28142|J01D... 4.0 "Samson Medical... "'cef-'" "1987" "cephalosporins... "'cef-(cef-)'" "1996" "J01DE01" "J01DE - Fourth... "J01D - OTHER B... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "None" "1:Synthetic Sm... 1 0 "Single Stereoi... 0 0 1 0 0 "Prescription O... "CO/N=C(\C(=O)N...
"CHEMBL222645" "FLOXACILLIN" "ABBOFLOX|BRL 2... "BRL 2039|BRL-2... 4.0 "" "'-cillin'" "1972" "penicillins" "'-cillin(-cill... "" "J01CF05" "J01CF - Beta-l... "J01C - BETA-LA... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "None" "1:Synthetic Sm... 1 0 "Single Stereoi... 0 0 0 0 0 "Unknown" "Cc1onc(-c2c(F)...
"CHEMBL2364632" "SARECYCLINE" "P-005672|P0056... "P-005672|P0056... 4.0 "Almirall Llc" "'-cycline'" "2012" "antibiotics (t... "'-cycline(-cyc... "2018" "J01AA14" "J01AA - Tetrac... "J01A - TETRACY... "J01 - ANTIBACT... "J - ANTIINFECT... "" "US-8318706-B2" "1:Synthetic Sm... 1 0 "Single Stereoi... 0 1 0 0 0 "Prescription O... "CON(C)Cc1ccc(O...
"CHEMBL62193" "SULFADIMETHOXI... "Madribon|NSC-6... "NSC-683544|NSC... 4.0 "" "'sulfa-'" "" "antimicrobials... "'sulfa-(sulfa-... "" "J01ED01" "J01ED - Long-a... "J01E - SULFONA... "J01 - ANTIBACT... "J - ANTIINFECT... "" "None" "1:Synthetic Sm... 1 0 "Achiral Molecu... 0 0 0 0 0 "Withdrawn" "COc1cc(NS(=O)(...
"CHEMBL2303613" "CEFODIZIME" "CEFODIZIME|CEF... "HR 221|HR 221 ... 4.0 "" "'cef-'" "" "cephalosporins... "'cef-(cef-)'" "" "J01DD09" "J01DD - Third-... "J01D - OTHER B... "J01 - ANTIBACT... "J - ANTIINFECT... "" "None" "1:Synthetic Sm... 0 0 "Single Stereoi... 0 0 0 0 0 "Unknown" "CO/N=C(\C(=O)N...
"CHEMBL31" "GATIFLOXACIN" "AM-1155|BMS-20... "AM-1155|BMS-20... 4.0 "Apotex Inc|Bri... "'-oxacin'" "1997" "antibacterials... "'-oxacin(-oxac... "1999" "S01AE06 | J01M... "S01AE - Fluoro... "S01A - ANTIINF... "S01 - OPHTHALM... "S - SENSORY OR... "Antibacterial" "None" "1:Synthetic Sm... 1 0 "Racemic Mixtur... 0 1 1 1 0 "Prescription O... "COc1c(N2CCNC(C...
"CHEMBL268869" "SULFAMETHOXYPY... "CL-13494|DEPOV... "CL-13494|NSC-7... 4.0 "" "'sulfa-'" "" "antimicrobials... "'sulfa-(sulfa-... "" "J01ED05" "J01ED - Long-a... "J01E - SULFONA... "J01 - ANTIBACT... "J - ANTIINFECT... "" "None" "1:Synthetic Sm... 1 0 "Achiral Molecu... 0 0 0 0 0 "Withdrawn" "COc1ccc(NS(=O)...
"CHEMBL3039597" "GENTAMICIN" "GENTAMICIN|GEN... "SCH-9724" 4.0 "Schering Corp ... "'-micin'" "1963" "antibiotics (M... "'-micin(-micin... "1970" "S01AA11 | S02A... "S01AA - Antibi... "S01A - ANTIINF... "S01 - OPHTHALM... "S - SENSORY OR... "Antibacterial" "None" "1:Synthetic Sm... 0 0 "Unknown" 0 0 1 1 1 "Prescription O... "CNC(C)[C@@H]1C...
"CHEMBL3182343" "PIVAMPICILLIN" "MK-191|PIVAMPI... "MK-191" 4.0 "" "'-cillin'" "1970" "penicillins" "'-cillin(-cill... "" "J01CA02" "J01CA - Penici... "J01C - BETA-LA... "J01 - ANTIBACT... "J - ANTIINFECT... "" "None" "1:Synthetic Sm... 1 0 "Single Stereoi... 1 0 0 0 0 "Unknown" "CC(C)(C)C(=O)O...
"CHEMBL2146161" "MIKAMYCIN" "MIKAMYCIN|PRIS... "RP 7293|RP-729... 4.0 "" "'-mycin'" "" "antibiotics (S... "'-mycin(-mycin... "" "J01FG01" "J01FG - Strept... "J01F - MACROLI... "J01 - ANTIBACT... "J - ANTIINFECT... "" "None" "1:Synthetic Sm... 0 0 "Unknown" 0 0 0 0 0 "Unknown" "CC1=C\[C@@H](O...
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
"CHEMBL572" "NITROFURANTOIN... "BERKFURIN|CEDU... "NSC-2107|NSC-4... 4.0 "Sun Pharmaceut... "'-toin'" "" "antiepileptics... "'-toin(-toin)'... "1953" "J01XE01 | J01X... "J01XE - Nitrof... "J01X - OTHER A... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial ... "None" "1:Synthetic Sm... 1 0 "Achiral Molecu... 0 1 1 0 0 "Prescription O... "O=C1CN(/N=C/c2...
"CHEMBL1596" "CARBENICILLIN ... "CARBENICILLIN ... "" 4.0 "Pfizer Laborat... "'-cillin'" "1972" "penicillins" "'-cillin(-cill... "1972" "J01CA05" "J01CA - Penici... "J01C - BETA-LA... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "None" "1:Synthetic Sm... 1 0 "Racemic Mixtur... 1 1 0 0 0 "Discontinued" "CC1(C)S[C@@H]2...
"CHEMBL44354" "CEFTAZIDIME" "CEFTAZIDIME|CE... "GR 20263|GR-20... 4.0 "Glaxosmithklin... "'cef-'" "1980" "cephalosporins... "'cef-(cef-)'" "1985" "J01DD02" "J01DD - Third-... "J01D - OTHER B... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "US-7112592-B2" "1:Synthetic Sm... 0 0 "Single Stereoi... 0 0 1 0 1 "Prescription O... "CC(C)(O/N=C(\C...
"CHEMBL29" "BENZYLPENICILL... "BENZYL PENICIL... "J01CE01|NSC-19... 4.0 "Hq Specialty P... "'-cillin'" "" "penicillins" "'-cillin(-cill... "1947" "J01CE01 | S01A... "J01CE - Beta-l... "J01C - BETA-LA... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "None" "1:Synthetic Sm... 1 0 "Single Stereoi... 0 1 1 0 1 "Prescription O... "CC1(C)S[C@@H]2...
"CHEMBL277100" "TEMAFLOXACIN" "Omniflox|TEMAF... "" 4.0 "" "'-oxacin'" "1988" "antibacterials... "'-oxacin(-oxac... "1992" "J01MA05" "J01MA - Fluoro... "J01M - QUINOLO... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial ... "None" "1:Synthetic Sm... 1 0 "Racemic Mixtur... 0 0 0 0 0 "Withdrawn" "CC1CN(c2cc3c(c...
"CHEMBL1731" "MEZLOCILLIN" "BAY-F-1353|BAY... "BAY-F-1353" 4.0 "Bayer Pharmace... "'-cillin'" "1976" "penicillins" "'-cillin(-cill... "1981" "J01CA10" "J01CA - Penici... "J01C - BETA-LA... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "None" "1:Synthetic Sm... 0 0 "Single Stereoi... 0 0 1 0 0 "Discontinued" "CC1(C)S[C@@H]2...
"CHEMBL2105612" "PROPICILLIN" "(1-PHENOXYPROP... "" 4.0 "" "'-cillin'" "" "penicillins" "'-cillin(-cill... "" "J01CE03" "J01CE - Beta-l... "J01C - BETA-LA... "J01 - ANTIBACT... "J - ANTIINFECT... "" "None" "1:Synthetic Sm... 1 0 "Racemic Mixtur... 0 0 0 0 0 "Unknown" "CCC(Oc1ccccc1)...
"CHEMBL8" "CIPROFLOXACIN" "BAY O 9867 FRE... "BAY O 9867 FRE... 4.0 "Inforlife Sa|A... "'-oxacin'" "1987" "antibacterials... "'-oxacin(-oxac... "1987" "J01MA02 | S03A... "J01MA - Fluoro... "J01M - QUINOLO... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "US-8318817-B2" "1:Synthetic Sm... 1 0 "Achiral Molecu... 0 1 1 1 1 "Prescription O... "O=C(O)c1cn(C2C...
"CHEMBL9" "NORFLOXACIN" "Baccidal|CHIBR... "MK-366|NSC-757... 4.0 "Merck Research... "'-oxacin'" "1984" "antibacterials... "'-oxacin(-oxac... "1986" "J01MA06 | S01A... "J01MA - Fluoro... "J01M - QUINOLO... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "None" "1:Synthetic Sm... 1 0 "Achiral Molecu... 0 1 0 1 1 "Prescription O... "CCn1cc(C(=O)O)...
"CHEMBL21" "SULFANILAMIDE" "ANILINE-P-SULF... "NSC-7618" 4.0 "Mylan Specialt... "'sulfa-'" "" "antimicrobials... "'sulfa-(sulfa-... "1985" "D06BA05 | J01E... "D06BA - Sulfon... "D06B - CHEMOTH... "D06 - ANTIBIOT... "D - DERMATOLOG... "" "None" "1:Synthetic Sm... 1 0 "Achiral Molecu... 0 0 0 1 0 "Prescription O... "Nc1ccc(S(N)(=O...
"CHEMBL4" "OFLOXACIN" "DL-8280|EXOCIN... "DL-8280|HOE 28... 4.0 "Bausch And Lom... "'-oxacin'" "1984" "antibacterials... "'-oxacin(-oxac... "1990" "J01MA01 | S02A... "J01MA - Fluoro... "J01M - QUINOLO... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "None" "1:Synthetic Sm... 1 0 "Racemic Mixtur... 0 1 1 1 1 "Prescription O... "CC1COc2c(N3CCN...
"CHEMBL530" "AMDINOCILLIN" "AMDINOCILLIN|C... "FL 1060|FL-106... 4.0 "Hoffmann La Ro... "'-cillin'" "1981" "penicillins" "'-cillin(-cill... "1984" "J01CA11" "J01CA - Penici... "J01C - BETA-LA... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "None" "1:Synthetic Sm... 1 0 "Single Stereoi... 0 0 1 0 0 "Discontinued" "CC1(C)S[C@@H]2... 
# Uncomment below if requiring a quick overview on all column names, 
# first ten variables in each column and each column data type
#print(df_ai.glimpse())

Under the “Availability Type” column, there were a few different availabilities for each anti-bacterial such as, “Discontinued”, “Withdrawn”, “Unknown” and “Prescription Only”.

df_ai.groupby("Availability Type").count()
shape: (4, 2)
Availability Type count
str u32
"Discontinued" 36
"Withdrawn" 7
"Unknown" 29
"Prescription O... 72

I only wanted to choose the “Prescription Only” ones, so the following filter condition was applied to the dataframe.

df_ai_rx = df_ai.filter(pl.col("Availability Type") == "Prescription Only")
df_ai_rx.head()
shape: (5, 29)
Parent Molecule Name Synonyms Research Codes Phase Drug Applicants USAN Stem USAN Year USAN Definition USAN Stem - Substem First Approval ATC Codes Level 4 ATC Codes Level 3 ATC Codes Level 2 ATC Codes Level 1 ATC Codes Indication Class Patent Drug Type Passes Rule of Five First In Class Chirality Prodrug Oral Parenteral Topical Black Box Availability Type Smiles
str str str str f64 str str str str str str str str str str str str str str i64 i64 str i64 i64 i64 i64 i64 str str
"CHEMBL186" "CEFEPIME" "BMY-28142|CEFE... "BMY-28142|J01D... 4.0 "Samson Medical... "'cef-'" "1987" "cephalosporins... "'cef-(cef-)'" "1996" "J01DE01" "J01DE - Fourth... "J01D - OTHER B... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "None" "1:Synthetic Sm... 1 0 "Single Stereoi... 0 0 1 0 0 "Prescription O... "CO/N=C(\C(=O)N...
"CHEMBL2364632" "SARECYCLINE" "P-005672|P0056... "P-005672|P0056... 4.0 "Almirall Llc" "'-cycline'" "2012" "antibiotics (t... "'-cycline(-cyc... "2018" "J01AA14" "J01AA - Tetrac... "J01A - TETRACY... "J01 - ANTIBACT... "J - ANTIINFECT... "" "US-8318706-B2" "1:Synthetic Sm... 1 0 "Single Stereoi... 0 1 0 0 0 "Prescription O... "CON(C)Cc1ccc(O...
"CHEMBL31" "GATIFLOXACIN" "AM-1155|BMS-20... "AM-1155|BMS-20... 4.0 "Apotex Inc|Bri... "'-oxacin'" "1997" "antibacterials... "'-oxacin(-oxac... "1999" "S01AE06 | J01M... "S01AE - Fluoro... "S01A - ANTIINF... "S01 - OPHTHALM... "S - SENSORY OR... "Antibacterial" "None" "1:Synthetic Sm... 1 0 "Racemic Mixtur... 0 1 1 1 0 "Prescription O... "COc1c(N2CCNC(C...
"CHEMBL3039597" "GENTAMICIN" "GENTAMICIN|GEN... "SCH-9724" 4.0 "Schering Corp ... "'-micin'" "1963" "antibiotics (M... "'-micin(-micin... "1970" "S01AA11 | S02A... "S01AA - Antibi... "S01A - ANTIINF... "S01 - OPHTHALM... "S - SENSORY OR... "Antibacterial" "None" "1:Synthetic Sm... 0 0 "Unknown" 0 0 1 1 1 "Prescription O... "CNC(C)[C@@H]1C...
"CHEMBL893" "DICLOXACILLIN" "BRL-1702|DICLO... "BRL-1702|R-134... 4.0 "Teva Pharmaceu... "'-cillin'" "1965" "penicillins" "'-cillin(-cill... "1968" "J01CF01" "J01CF - Beta-l... "J01C - BETA-LA... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "None" "1:Synthetic Sm... 1 0 "Single Stereoi... 0 1 0 0 0 "Prescription O... "Cc1onc(-c2c(Cl...

In preparation for possible future work on building machine learning models on this line of work, I looked into Datamol’s function on pre-processing molecules (shown in the next section), as it involved converting SMILES strings into SELFIES, which were considered to be more robust than SMILES.

However, I kept running into an error, with the error message showing the SMILES column was empty. After a few tries I realised that I’ve actually forgotten to check whether there were any missing SMILES in the column. So here I’ve filtered the SMILES column to look for any missing SMILES

df_ai_rx.filter(pl.col("Smiles") == "")
shape: (1, 29)
Parent Molecule Name Synonyms Research Codes Phase Drug Applicants USAN Stem USAN Year USAN Definition USAN Stem - Substem First Approval ATC Codes Level 4 ATC Codes Level 3 ATC Codes Level 2 ATC Codes Level 1 ATC Codes Indication Class Patent Drug Type Passes Rule of Five First In Class Chirality Prodrug Oral Parenteral Topical Black Box Availability Type Smiles
str str str str f64 str str str str str str str str str str str str str str i64 i64 str i64 i64 i64 i64 i64 str str
"CHEMBL3989751" "NEOMYCIN" "FRADIOMYCIN|KA... "" 4.0 "Bayer Pharmace... "'-mycin'" "1966" "antibiotics (S... "'-mycin(-mycin... "1957" "R02AB01 | S01A... "R02AB - Antibi... "R02A - THROAT ... "R02 - THROAT P... "R - RESPIRATOR... "" "None" "1:Synthetic Sm... 0 0 "Single Stereoi... 0 1 0 1 1 "Prescription O... ""

Neomycin was the only compound found to have no SMILES recorded. To fix this error, I then used the “when-then-otherwise” expression in Polars again (used in previous post) to replace the empty string in the dataframe. A code example below was kindly adapted from StackOverflow from this link, and code example as shown below.

```{python}
only_these = ['str1', 'str2']
df.with_columns([
    pl.when(pl.col(only_these).str.lengths() == 0)
    .then(None)
    .otherwise(pl.col(only_these))
    .keep_name()
])
```

This was what I’ve done to amend the issue.

# Canonical SMILES for neomycin was extracted from PubChem 
# (https://pubchem.ncbi.nlm.nih.gov/compound/Neomycin)

df_ai_rx = df_ai_rx.with_columns([
    pl.when(pl.col("Smiles").str.lengths() == 0)
    .then("C1C(C(C(C(C1N)OC2C(C(C(C(O2)CN)O)O)N)OC3C(C(C(O3)CO)OC4C(C(C(C(O4)CN)O)O)N)O)O)N")
    .otherwise(pl.col("Smiles"))
    .keep_name()
])

df_ai_rx
shape: (72, 29)
Parent Molecule Name Synonyms Research Codes Phase Drug Applicants USAN Stem USAN Year USAN Definition USAN Stem - Substem First Approval ATC Codes Level 4 ATC Codes Level 3 ATC Codes Level 2 ATC Codes Level 1 ATC Codes Indication Class Patent Drug Type Passes Rule of Five First In Class Chirality Prodrug Oral Parenteral Topical Black Box Availability Type Smiles
str str str str f64 str str str str str str str str str str str str str str i64 i64 str i64 i64 i64 i64 i64 str str
"CHEMBL186" "CEFEPIME" "BMY-28142|CEFE... "BMY-28142|J01D... 4.0 "Samson Medical... "'cef-'" "1987" "cephalosporins... "'cef-(cef-)'" "1996" "J01DE01" "J01DE - Fourth... "J01D - OTHER B... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "None" "1:Synthetic Sm... 1 0 "Single Stereoi... 0 0 1 0 0 "Prescription O... "CO/N=C(\C(=O)N...
"CHEMBL2364632" "SARECYCLINE" "P-005672|P0056... "P-005672|P0056... 4.0 "Almirall Llc" "'-cycline'" "2012" "antibiotics (t... "'-cycline(-cyc... "2018" "J01AA14" "J01AA - Tetrac... "J01A - TETRACY... "J01 - ANTIBACT... "J - ANTIINFECT... "" "US-8318706-B2" "1:Synthetic Sm... 1 0 "Single Stereoi... 0 1 0 0 0 "Prescription O... "CON(C)Cc1ccc(O...
"CHEMBL31" "GATIFLOXACIN" "AM-1155|BMS-20... "AM-1155|BMS-20... 4.0 "Apotex Inc|Bri... "'-oxacin'" "1997" "antibacterials... "'-oxacin(-oxac... "1999" "S01AE06 | J01M... "S01AE - Fluoro... "S01A - ANTIINF... "S01 - OPHTHALM... "S - SENSORY OR... "Antibacterial" "None" "1:Synthetic Sm... 1 0 "Racemic Mixtur... 0 1 1 1 0 "Prescription O... "COc1c(N2CCNC(C...
"CHEMBL3039597" "GENTAMICIN" "GENTAMICIN|GEN... "SCH-9724" 4.0 "Schering Corp ... "'-micin'" "1963" "antibiotics (M... "'-micin(-micin... "1970" "S01AA11 | S02A... "S01AA - Antibi... "S01A - ANTIINF... "S01 - OPHTHALM... "S - SENSORY OR... "Antibacterial" "None" "1:Synthetic Sm... 0 0 "Unknown" 0 0 1 1 1 "Prescription O... "CNC(C)[C@@H]1C...
"CHEMBL893" "DICLOXACILLIN" "BRL-1702|DICLO... "BRL-1702|R-134... 4.0 "Teva Pharmaceu... "'-cillin'" "1965" "penicillins" "'-cillin(-cill... "1968" "J01CF01" "J01CF - Beta-l... "J01C - BETA-LA... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "None" "1:Synthetic Sm... 1 0 "Single Stereoi... 0 1 0 0 0 "Prescription O... "Cc1onc(-c2c(Cl...
"CHEMBL1449" "TICARCILLIN" "TICARCILLIN|Ti... "" 4.0 "Glaxosmithklin... "'-cillin'" "1973" "penicillins" "'-cillin(-cill... "1976" "J01CA13" "J01CA - Penici... "J01C - BETA-LA... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "None" "1:Synthetic Sm... 1 0 "Single Stereoi... 0 0 1 0 0 "Prescription O... "CC1(C)S[C@@H]2...
"CHEMBL1220" "TINIDAZOLE" "CP-12,574|CP-1... "CP-12,574|CP-1... 4.0 "Mission Pharma... "'-nidazole'" "1970" "antiprotozoal ... "'-nidazole(-ni... "2004" "P01AB02 | J01X... "P01AB - Nitroi... "P01A - AGENTS ... "P01 - ANTIPROT... "P - ANTIPARASI... "Antiprotozoal" "None" "1:Synthetic Sm... 1 0 "Achiral Molecu... 1 1 0 0 1 "Prescription O... "CCS(=O)(=O)CCn...
"CHEMBL501122" "CEFTAROLINE FO... "CEFTAROLINE FO... "PPI-0903|TAK 5... 4.0 "Apotex Inc|All... "'cef-; fos-'" "2006" "cephalosporins... "'cef-(cef-); f... "2010" "J01DI02" "J01DI - Other ... "J01D - OTHER B... "J01 - ANTIBACT... "J - ANTIINFECT... "" "US-6417175-B1" "1:Synthetic Sm... 0 0 "Single Stereoi... 1 0 1 0 0 "Prescription O... "CCO/N=C(\C(=O)...
"CHEMBL137" "METRONIDAZOLE" "ACEA|ANABACT|B... "BAY-5360|BAYER... 4.0 "Inforlife Sa|L... "'-nidazole'" "1962" "antiprotozoal ... "'-nidazole(-ni... "1963" "D06BX01 | P01A... "D06BX - Other ... "D06B - CHEMOTH... "D06 - ANTIBIOT... "D - DERMATOLOG... "Antibacterial,... "US-6881726-B2" "1:Synthetic Sm... 1 0 "Achiral Molecu... 0 1 1 1 1 "Prescription O... "Cc1ncc([N+](=O...
"CHEMBL376140" "TIGECYCLINE" "TIGECYCLINE|Ty... "WAY-GAR-936" 4.0 "Apotex Inc|Fre... "'-cycline'" "2002" "antibiotics (t... "'-cycline(-cyc... "2005" "J01AA12" "J01AA - Tetrac... "J01A - TETRACY... "J01 - ANTIBACT... "J - ANTIINFECT... "" "US-7879828-B2" "1:Synthetic Sm... 0 0 "Single Stereoi... 0 0 1 0 1 "Prescription O... "CN(C)c1cc(NC(=...
"CHEMBL1741" "CLARITHROMYCIN... "6-O-METHYLERYT... "A-56268|ABBOTT... 4.0 "Sunshine Lake ... "'-mycin'" "1988" "antibiotics (S... "'-mycin(-mycin... "1991" "J01FA09" "J01FA - Macrol... "J01F - MACROLI... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "US-7977488-B2" "1:Synthetic Sm... 0 0 "Single Stereoi... 0 1 0 0 0 "Prescription O... "CC[C@H]1OC(=O)...
"CHEMBL1747" "TOBRAMYCIN" "47663|Aktob|BE... "47663|NSC-1805... 4.0 "Igi Laboratori... "'-mycin'" "1972" "antibiotics (S... "'-mycin(-mycin... "1975" "J01GB01 | S01A... "J01GB - Other ... "J01G - AMINOGL... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "US-6987094-B2" "1:Synthetic Sm... 0 0 "Single Stereoi... 0 0 1 1 1 "Prescription O... "NC[C@H]1O[C@H]...
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
"CHEMBL404" "TAZOBACTAM" "CL 298,741|CL-... "CL 298,741|CL-... 4.0 "Wytells Pharma... "'-bactam'" "1989" "beta-lactamase... "'-bactam(-bact... "1993" "J01CG02" "J01CG - Beta-l... "J01C - BETA-LA... "J01 - ANTIBACT... "J - ANTIINFECT... "Inhibitor (bet... "US-6900184-B2" "1:Synthetic Sm... 1 0 "Single Stereoi... 0 0 1 0 0 "Prescription O... "C[C@]1(Cn2ccnn...
"CHEMBL1689772" "OMADACYCLINE" "AMADACYCLINE|B... "BAY 73-6944|PT... 4.0 "Paratek Pharma... "'-cycline'" "2009" "antibiotics (t... "'-cycline(-cyc... "2018" "J01AA15" "J01AA - Tetrac... "J01A - TETRACY... "J01 - ANTIBACT... "J - ANTIINFECT... "" "US-7326696-B2" "1:Synthetic Sm... 0 0 "Single Stereoi... 0 1 1 0 0 "Prescription O... "CN(C)c1cc(CNCC...
"CHEMBL3989974" "CEFIDEROCOL" "CEFIDEROCOL|GS... "GSK2696266|S-6... 4.0 "Shionogi Inc" "'cef-'" "2017" "cephalosporins... "'cef-(cef-)'" "2019" "J01DI04" "J01DI - Other ... "J01D - OTHER B... "J01 - ANTIBACT... "J - ANTIINFECT... "" "US-9238657-B2" "1:Synthetic Sm... 0 0 "Single Stereoi... 0 0 1 0 0 "Prescription O... "CC(C)(O/N=C(\C...
"CHEMBL1435" "CEFAZOLIN" "CEFAZOLIN|CEPH... "J01DB04|SK&F-4... 4.0 "Glaxosmithklin... "'cef-'" "1972" "cephalosporins... "'cef-(cef-)'" "1973" "J01DB04" "J01DB - First-... "J01D - OTHER B... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial ... "None" "1:Synthetic Sm... 0 0 "Single Stereoi... 0 0 1 0 0 "Prescription O... "Cc1nnc(SCC2=C(...
"CHEMBL3989751" "NEOMYCIN" "FRADIOMYCIN|KA... "" 4.0 "Bayer Pharmace... "'-mycin'" "1966" "antibiotics (S... "'-mycin(-mycin... "1957" "R02AB01 | S01A... "R02AB - Antibi... "R02A - THROAT ... "R02 - THROAT P... "R - RESPIRATOR... "" "None" "1:Synthetic Sm... 0 0 "Single Stereoi... 0 1 0 1 1 "Prescription O... "C1C(C(C(C(C1N)...
"CHEMBL572" "NITROFURANTOIN... "BERKFURIN|CEDU... "NSC-2107|NSC-4... 4.0 "Sun Pharmaceut... "'-toin'" "" "antiepileptics... "'-toin(-toin)'... "1953" "J01XE01 | J01X... "J01XE - Nitrof... "J01X - OTHER A... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial ... "None" "1:Synthetic Sm... 1 0 "Achiral Molecu... 0 1 1 0 0 "Prescription O... "O=C1CN(/N=C/c2...
"CHEMBL44354" "CEFTAZIDIME" "CEFTAZIDIME|CE... "GR 20263|GR-20... 4.0 "Glaxosmithklin... "'cef-'" "1980" "cephalosporins... "'cef-(cef-)'" "1985" "J01DD02" "J01DD - Third-... "J01D - OTHER B... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "US-7112592-B2" "1:Synthetic Sm... 0 0 "Single Stereoi... 0 0 1 0 1 "Prescription O... "CC(C)(O/N=C(\C...
"CHEMBL29" "BENZYLPENICILL... "BENZYL PENICIL... "J01CE01|NSC-19... 4.0 "Hq Specialty P... "'-cillin'" "" "penicillins" "'-cillin(-cill... "1947" "J01CE01 | S01A... "J01CE - Beta-l... "J01C - BETA-LA... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "None" "1:Synthetic Sm... 1 0 "Single Stereoi... 0 1 1 0 1 "Prescription O... "CC1(C)S[C@@H]2...
"CHEMBL8" "CIPROFLOXACIN" "BAY O 9867 FRE... "BAY O 9867 FRE... 4.0 "Inforlife Sa|A... "'-oxacin'" "1987" "antibacterials... "'-oxacin(-oxac... "1987" "J01MA02 | S03A... "J01MA - Fluoro... "J01M - QUINOLO... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "US-8318817-B2" "1:Synthetic Sm... 1 0 "Achiral Molecu... 0 1 1 1 1 "Prescription O... "O=C(O)c1cn(C2C...
"CHEMBL9" "NORFLOXACIN" "Baccidal|CHIBR... "MK-366|NSC-757... 4.0 "Merck Research... "'-oxacin'" "1984" "antibacterials... "'-oxacin(-oxac... "1986" "J01MA06 | S01A... "J01MA - Fluoro... "J01M - QUINOLO... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "None" "1:Synthetic Sm... 1 0 "Achiral Molecu... 0 1 0 1 1 "Prescription O... "CCn1cc(C(=O)O)...
"CHEMBL21" "SULFANILAMIDE" "ANILINE-P-SULF... "NSC-7618" 4.0 "Mylan Specialt... "'sulfa-'" "" "antimicrobials... "'sulfa-(sulfa-... "1985" "D06BA05 | J01E... "D06BA - Sulfon... "D06B - CHEMOTH... "D06 - ANTIBIOT... "D - DERMATOLOG... "" "None" "1:Synthetic Sm... 1 0 "Achiral Molecu... 0 0 0 1 0 "Prescription O... "Nc1ccc(S(N)(=O...
"CHEMBL4" "OFLOXACIN" "DL-8280|EXOCIN... "DL-8280|HOE 28... 4.0 "Bausch And Lom... "'-oxacin'" "1984" "antibacterials... "'-oxacin(-oxac... "1990" "J01MA01 | S02A... "J01MA - Fluoro... "J01M - QUINOLO... "J01 - ANTIBACT... "J - ANTIINFECT... "Antibacterial" "None" "1:Synthetic Sm... 1 0 "Racemic Mixtur... 0 1 1 1 1 "Prescription O... "CC1COc2c(N3CCN... 
# Keeping only selected columns with information needed for later use
df_ai_rx = df_ai_rx.select(["Smiles", "Name", "USAN Definition", "Level 4 ATC Codes"])
df_ai_rx.head()
shape: (5, 4)
Smiles Name USAN Definition Level 4 ATC Codes
str str str str
"CO/N=C(\C(=O)N... "CEFEPIME" "cephalosporins... "J01DE - Fourth...
"CON(C)Cc1ccc(O... "SARECYCLINE" "antibiotics (t... "J01AA - Tetrac...
"COc1c(N2CCNC(C... "GATIFLOXACIN" "antibacterials... "S01AE - Fluoro...
"CNC(C)[C@@H]1C... "GENTAMICIN" "antibiotics (M... "S01AA - Antibi...
"Cc1onc(-c2c(Cl... "DICLOXACILLIN" "penicillins" "J01CF - Beta-l...

The “Smiles” column name was changed below to ensure _preprocess function would work since the parameter “smiles_column” in _preprocess function had “smiles” with lowercase “s” (this of course could be the other way round, where we could change the parameter name in the function instead - the column name and parameter name had to match for the function to work). The “Name” column was changed accordingly for similar reason.

df_ai_rx = df_ai_rx.rename({"Smiles": "smiles", "Name": "names"})
df_ai_rx.head()
shape: (5, 4)
smiles names USAN Definition Level 4 ATC Codes
str str str str
"CO/N=C(\C(=O)N... "CEFEPIME" "cephalosporins... "J01DE - Fourth...
"CON(C)Cc1ccc(O... "SARECYCLINE" "antibiotics (t... "J01AA - Tetrac...
"COc1c(N2CCNC(C... "GATIFLOXACIN" "antibacterials... "S01AE - Fluoro...
"CNC(C)[C@@H]1C... "GENTAMICIN" "antibiotics (M... "S01AA - Antibi...
"Cc1onc(-c2c(Cl... "DICLOXACILLIN" "penicillins" "J01CF - Beta-l...

I also wanted to change the all capitalised compound names into lowercases for the ease of reading.

# Convert all compounds to lowercases
df_ai_rx = df_ai_rx.with_columns(pl.col("names").str.to_lowercase())
df_ai_rx.head()
shape: (5, 4)
smiles names USAN Definition Level 4 ATC Codes
str str str str
"CO/N=C(\C(=O)N... "cefepime" "cephalosporins... "J01DE - Fourth...
"CON(C)Cc1ccc(O... "sarecycline" "antibiotics (t... "J01AA - Tetrac...
"COc1c(N2CCNC(C... "gatifloxacin" "antibacterials... "S01AE - Fluoro...
"CNC(C)[C@@H]1C... "gentamicin" "antibiotics (M... "S01AA - Antibi...
"Cc1onc(-c2c(Cl... "dicloxacillin" "penicillins" "J01CF - Beta-l...

Since Datamol was built as a thin layer library on top of RDKit, which was really only compatible with Pandas, I added the following step to convert the dataframe into a Pandas one.

df_ai_pd = df_ai_rx.to_pandas()
df_ai_pd.head()
smiles names USAN Definition Level 4 ATC Codes
0 CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N... cefepime cephalosporins J01DE - Fourth-generation cephalosporins
1 CON(C)Cc1ccc(O)c2c1C[C@H]1C[C@H]3[C@H](N(C)C)C... sarecycline antibiotics (tetracycline derivatives) J01AA - Tetracyclines
2 COc1c(N2CCNC(C)C2)c(F)cc2c(=O)c(C(=O)O)cn(C3CC... gatifloxacin antibacterials (quinolone derivatives) S01AE - Fluoroquinolones | J01MA - Fluoroquino...
3 CNC(C)[C@@H]1CC[C@@H](N)[C@@H](O[C@H]2[C@H](O)... gentamicin antibiotics (Micromonospora strains) S01AA - Antibiotics | S02AA - Antiinfectives |...
4 Cc1onc(-c2c(Cl)cccc2Cl)c1C(=O)N[C@@H]1C(=O)N2[... dicloxacillin penicillins J01CF - Beta-lactamase resistant penicillins 
# Check the dataframe has been converted from Polars to Pandas
type(df_ai_pd)
pandas.core.frame.DataFrame


Pre-processing and standardising molecules

I have borrowed and adapted the _preprocess function from Datamol (link here), as shown below. One of the convenient features in this function was that it also included a conversion from “mol” (RDKit molecule) to SELFIES amongst several other common molecular representations such as InChI3 and SMILES.

# Pre-process molecules using _preprocess function - adapted from datamol.io

smiles_column = "smiles"

dm.disable_rdkit_log()

def _preprocess(row):
    mol = dm.to_mol(row[smiles_column], ordered=True)
    mol = dm.fix_mol(mol)
    mol = dm.sanitize_mol(mol, sanifix=True, charge_neutral=False)
    mol = dm.standardize_mol(
        mol,
        disconnect_metals=False,
        normalize=True,
        reionize=True,
        uncharge=False,
        stereo=True,
    )

    row["standard_smiles"] = dm.standardize_smiles(dm.to_smiles(mol))
    row["selfies"] = dm.to_selfies(mol)
    row["inchi"] = dm.to_inchi(mol)
    row["inchikey"] = dm.to_inchikey(mol)
    return row


Converting multiple SMILES into multiple SELFIES

There were two ways to convert multiple SMILES into multiple SELFIES (there might be more options, but I’ve found these two for now):

# Method one - using lambda function:
# Uncomment code below to run
# df_ai_pd["selfies"] = df_ai_pd["Smiles"].apply(lambda x: dm.to_selfies(x))
# df_ai_pd

# Method two - using _preprocess function:
data_mol_clean = df_ai_pd.apply(_preprocess, axis = 1)
data_mol_clean
smiles names USAN Definition Level 4 ATC Codes standard_smiles selfies inchi inchikey
0 CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N... cefepime cephalosporins J01DE - Fourth-generation cephalosporins CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N... [C][O][/N][=C][Branch2][Ring2][#Branch2][\C][=... InChI=1S/C19H24N6O5S2/c1-25(5-3-4-6-25)7-10-8-... HVFLCNVBZFFHBT-ZKDACBOMSA-N
1 CON(C)Cc1ccc(O)c2c1C[C@H]1C[C@H]3[C@H](N(C)C)C... sarecycline antibiotics (tetracycline derivatives) J01AA - Tetracyclines CON(C)Cc1ccc(O)c2c1C[C@H]1C[C@H]3[C@H](N(C)C)C... [C][O][N][Branch1][C][C][C][C][=C][C][=C][Bran... InChI=1S/C24H29N3O8/c1-26(2)18-13-8-11-7-12-10... PQJQFLNBMSCUSH-SBAJWEJLSA-N
2 COc1c(N2CCNC(C)C2)c(F)cc2c(=O)c(C(=O)O)cn(C3CC... gatifloxacin antibacterials (quinolone derivatives) S01AE - Fluoroquinolones | J01MA - Fluoroquino... COc1c(N2CCNC(C)C2)c(F)cc2c(=O)c(C(=O)O)cn(C3CC... [C][O][C][=C][Branch1][N][N][C][C][N][C][Branc... InChI=1S/C19H22FN3O4/c1-10-8-22(6-5-21-10)16-1... XUBOMFCQGDBHNK-UHFFFAOYSA-N
3 CNC(C)[C@@H]1CC[C@@H](N)[C@@H](O[C@H]2[C@H](O)... gentamicin antibiotics (Micromonospora strains) S01AA - Antibiotics | S02AA - Antiinfectives |... CNC(C)[C@@H]1CC[C@@H](N)[C@@H](O[C@H]2[C@H](O)... [C][N][C][Branch1][C][C][C@@H1][C][C][C@@H1][B... InChI=1S/C21H43N5O7.C20H41N5O7.C19H39N5O7/c1-9... NPEFREDMMVQEPL-RWPARATISA-N
4 Cc1onc(-c2c(Cl)cccc2Cl)c1C(=O)N[C@@H]1C(=O)N2[... dicloxacillin penicillins J01CF - Beta-lactamase resistant penicillins Cc1onc(-c2c(Cl)cccc2Cl)c1C(=O)N[C@@H]1C(=O)N2[... [C][C][O][N][=C][Branch1][=N][C][=C][Branch1][... InChI=1S/C19H17Cl2N3O5S/c1-7-10(12(23-29-7)11-... YFAGHNZHGGCZAX-JKIFEVAISA-N
... ... ... ... ... ... ... ... ...
67 CC1(C)S[C@@H]2[C@H](NC(=O)Cc3ccccc3)C(=O)N2[C@... benzylpenicillin penicillins J01CE - Beta-lactamase sensitive penicillins |... CC1(C)S[C@@H]2[C@H](NC(=O)Cc3ccccc3)C(=O)N2[C@... [C][C][Branch1][C][C][S][C@@H1][C@H1][Branch1]... InChI=1S/C16H18N2O4S/c1-16(2)12(15(21)22)18-13... JGSARLDLIJGVTE-MBNYWOFBSA-N
68 O=C(O)c1cn(C2CC2)c2cc(N3CCNCC3)c(F)cc2c1=O ciprofloxacin antibacterials (quinolone derivatives) J01MA - Fluoroquinolones | S03AA - Antiinfecti... O=C(O)c1cn(C2CC2)c2cc(N3CCNCC3)c(F)cc2c1=O [O][=C][Branch1][C][O][C][=C][N][Branch1][=Bra... InChI=1S/C17H18FN3O3/c18-13-7-11-14(8-15(13)20... MYSWGUAQZAJSOK-UHFFFAOYSA-N
69 CCn1cc(C(=O)O)c(=O)c2cc(F)c(N3CCNCC3)cc21 norfloxacin antibacterials (quinolone derivatives) J01MA - Fluoroquinolones | S01AE - Fluoroquino... CCn1cc(C(=O)O)c(=O)c2cc(F)c(N3CCNCC3)cc21 [C][C][N][C][=C][Branch1][=Branch1][C][=Branch... InChI=1S/C16H18FN3O3/c1-2-19-9-11(16(22)23)15(... OGJPXUAPXNRGGI-UHFFFAOYSA-N
70 Nc1ccc(S(N)(=O)=O)cc1 sulfanilamide antimicrobials (sulfonamides derivatives) D06BA - Sulfonamides | J01EB - Short-acting su... Nc1ccc(S(N)(=O)=O)cc1 [N][C][=C][C][=C][Branch1][=Branch2][S][Branch... InChI=1S/C6H8N2O2S/c7-5-1-3-6(4-2-5)11(8,9)10/... FDDDEECHVMSUSB-UHFFFAOYSA-N
71 CC1COc2c(N3CCN(C)CC3)c(F)cc3c(=O)c(C(=O)O)cn1c23 ofloxacin antibacterials (quinolone derivatives) J01MA - Fluoroquinolones | S02AA - Antiinfecti... CC1COc2c(N3CCN(C)CC3)c(F)cc3c(=O)c(C(=O)O)cn1c23 [C][C][C][O][C][=C][Branch1][N][N][C][C][N][Br... InChI=1S/C18H20FN3O4/c1-10-9-26-17-14-11(16(23... GSDSWSVVBLHKDQ-UHFFFAOYSA-N

72 rows × 8 columns

Converting a single SMILES into a SELFIES

To convert only one SMILES string into a SELFIES, the following code example should work with Datamol.

```{python}
selfies = dm.to_selfies("O=C(N[C@H](CO)[C@H](O)c1ccc([N+](=O)[O-])cc1)C(Cl)Cl")
selfies
```


Visualise compounds in 2D using Datamol

The images generated below might be quite small to see or read clearly. I’ve tried to increase the molecule size (mol_size) and also reduce the column numbers, but it still appeared the same. However, if the code was run in say VS Code, the compound images would appear larger when increasing the mol_size.

# Grab all SMILES of the cleaned/pre-processed ChEMBL anti-infectives
df_ai_sm = data_mol_clean["standard_smiles"]

# Load a list of these molecules in SMILES
# dm.to_mol() has sanitize = True set as default
mol_ls = [dm.to_mol(smile) for smile in df_ai_sm]

# Alternative way to convert dataframe into a list of mols (same as mol_ls)
# mols = dm.from_df(df_name, smiles_column = "Smiles")

# Add compound name for each 2D image
legends_c = list(data_mol_clean["names"])

# Convert the list of molecules into 2D images
dm.to_image(mol_ls, n_cols = 4, mol_size = (400, 400), legends = legends_c)


Extract scaffolds
# Extract Murcko scaffolds from mol_ls (ChEMBL anti-infectives)
m_scaffolds = [dm.to_scaffold_murcko(mol) for mol in mol_ls]
dm.to_image(m_scaffolds, mol_size = (400, 400), legends = legends_c)


Filamenting temperature-sensitive mutant Z (FtsZ) compounds

Data cleaning

This section focuses on 3 compounds from this paper: Lin, H.-Y.J.; Battaje, R.R.; Tan, J.; Doddareddy, M.; Dhaked, H.P.S.; Srivastava, S.; Hawkins, B.A.; Al-Shdifat, L.M.H.; Hibbs, D.E.; Panda, D.; et al. Discovery of 2’,6-Bis(4-hydroxybenzyl)-2-acetylcyclohexanone, a Novel FtsZ Inhibitor. Molecules 2022, 27, 6993. Obviously from the authorship shown, this was from my PhD work and out of interests, I just wanted to look a bit further into them and compare with known anti-infectives.

Before I started cleaning any data on FtsZ compounds, I found this useful website, OPSIN: Open Parser for Systematic IUPAC nomenclature, with this link to the journal paper as an acknowledgement of the work. I’ve managed to convert these 3 FtsZ compounds by using their IUPAC names, which were inputted into OPSIN, and converted into the corresponding InChI or SMILES strings.

After that, I started by converting the InChI of compound 1 into a RDKit molecule, which could be visualised in 2D below.

# Convert compound 1 to mol from InChI
cpd1 = dm.from_inchi("InChI=1S/C22H20O4/c23-18-9-4-15(5-10-18)8-13-21(25)20-3-1-2-17(22(20)26)14-16-6-11-19(24)12-7-16/h4-14,20,23-24H,1-3H2/b13-8+,17-14+")
cpd1

I then converted compound 2 using SMILES string instead.

# Convert compound 2 SMILES to mol
cpd2 = dm.to_mol("OC1=C(C=CC=C1CC1=CC=C(C=C1)O)C(CCC1=CC=C(C=C1)O)=O")
cpd2

Same thing followed for compound 3.

# Convert compound 3 SMILES to mol
cpd3 = dm.to_mol("OC1=CC=C(CC2C(C(CCC2)C(CCC2=CC=C(C=C2)O)=O)=O)C=C1")
cpd3

# Save these 3 compounds into a list
mol_lst = [cpd1, cpd2, cpd3]
mol_lst
[<rdkit.Chem.rdchem.Mol at 0x13426b7d0>,
 <rdkit.Chem.rdchem.Mol at 0x13426a500>,
 <rdkit.Chem.rdchem.Mol at 0x13426ba70>]
# Convert a list of mols into a dataframe
df = dm.to_df(mol_lst)
df
smiles
0 O=C(/C=C/c1ccc(O)cc1)C1CCC/C(=C\c2ccc(O)cc2)C1=O
1 O=C(CCc1ccc(O)cc1)c1cccc(Cc2ccc(O)cc2)c1O
2 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O

Then the compound numbers were added into the dataframe as well.

names = ["Compound_1", "Compound_2", "Compound_3"]
df["names"] = names
df
smiles names
0 O=C(/C=C/c1ccc(O)cc1)C1CCC/C(=C\c2ccc(O)cc2)C1=O Compound_1
1 O=C(CCc1ccc(O)cc1)c1cccc(Cc2ccc(O)cc2)c1O Compound_2
2 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O Compound_3


Pre-processing and standardising molecules
# Pre-processing FtsZ compounds
data_cleaned = df.apply(_preprocess, axis=1)
data_cleaned
smiles names standard_smiles selfies inchi inchikey
0 O=C(/C=C/c1ccc(O)cc1)C1CCC/C(=C\c2ccc(O)cc2)C1=O Compound_1 O=C(/C=C/c1ccc(O)cc1)C1CCC/C(=C\c2ccc(O)cc2)C1=O [O][=C][Branch1][=C][/C][=C][/C][=C][C][=C][Br... InChI=1S/C22H20O4/c23-18-9-4-15(5-10-18)8-13-2... QNBFRAOWJNMPAF-ZIQQYUHESA-N
1 O=C(CCc1ccc(O)cc1)c1cccc(Cc2ccc(O)cc2)c1O Compound_2 O=C(CCc1ccc(O)cc1)c1cccc(Cc2ccc(O)cc2)c1O [O][=C][Branch1][=C][C][C][C][=C][C][=C][Branc... InChI=1S/C22H20O4/c23-18-9-4-15(5-10-18)8-13-2... CWVXMBYGWRWONE-UHFFFAOYSA-N
2 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O [O][=C][Branch1][=C][C][C][C][=C][C][=C][Branc... InChI=1S/C22H24O4/c23-18-9-4-15(5-10-18)8-13-2... KWPXNGBFYRHREW-UHFFFAOYSA-N


Visualise compounds in 2D using Datamol
# Grab all SMILES from cleaned FtsZ compound dataset
df_ai_ftsz = data_cleaned["standard_smiles"]

# Load a list of these molecules in SMILES
mol_ftsz_list = [dm.to_mol(smile) for smile in df_ai_ftsz]

# Add compound names for each 2D image of compounds
legends = list(data_cleaned["names"])

# Convert the list of molecules into 2D images
dm.to_image(mol_ftsz_list, n_cols = 5, mol_size = (400, 400), legends = legends)


Extract scaffolds
# Get Murcko scaffolds of FtsZ compounds
m_ftsz_scaffolds = [dm.to_scaffold_murcko(mol) for mol in mol_ftsz_list]
dm.to_image(m_ftsz_scaffolds, mol_size = (400, 400), legends = legends)


Combining ChEMBL anti-infectives and FtsZ compounds

Combining dataframes

In this part, I wanted to combine the two dataframes I had from above, since my next step was to compare the scaffolds between ChEMBL prescription-only anti-infectives and FtsZ compounds.

combined_lists = [data_mol_clean, data_cleaned]
full_data = pd.concat(combined_lists, ignore_index = True)
full_data.head()
smiles names USAN Definition Level 4 ATC Codes standard_smiles selfies inchi inchikey
0 CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N... cefepime cephalosporins J01DE - Fourth-generation cephalosporins CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N... [C][O][/N][=C][Branch2][Ring2][#Branch2][\C][=... InChI=1S/C19H24N6O5S2/c1-25(5-3-4-6-25)7-10-8-... HVFLCNVBZFFHBT-ZKDACBOMSA-N
1 CON(C)Cc1ccc(O)c2c1C[C@H]1C[C@H]3[C@H](N(C)C)C... sarecycline antibiotics (tetracycline derivatives) J01AA - Tetracyclines CON(C)Cc1ccc(O)c2c1C[C@H]1C[C@H]3[C@H](N(C)C)C... [C][O][N][Branch1][C][C][C][C][=C][C][=C][Bran... InChI=1S/C24H29N3O8/c1-26(2)18-13-8-11-7-12-10... PQJQFLNBMSCUSH-SBAJWEJLSA-N
2 COc1c(N2CCNC(C)C2)c(F)cc2c(=O)c(C(=O)O)cn(C3CC... gatifloxacin antibacterials (quinolone derivatives) S01AE - Fluoroquinolones | J01MA - Fluoroquino... COc1c(N2CCNC(C)C2)c(F)cc2c(=O)c(C(=O)O)cn(C3CC... [C][O][C][=C][Branch1][N][N][C][C][N][C][Branc... InChI=1S/C19H22FN3O4/c1-10-8-22(6-5-21-10)16-1... XUBOMFCQGDBHNK-UHFFFAOYSA-N
3 CNC(C)[C@@H]1CC[C@@H](N)[C@@H](O[C@H]2[C@H](O)... gentamicin antibiotics (Micromonospora strains) S01AA - Antibiotics | S02AA - Antiinfectives |... CNC(C)[C@@H]1CC[C@@H](N)[C@@H](O[C@H]2[C@H](O)... [C][N][C][Branch1][C][C][C@@H1][C][C][C@@H1][B... InChI=1S/C21H43N5O7.C20H41N5O7.C19H39N5O7/c1-9... NPEFREDMMVQEPL-RWPARATISA-N
4 Cc1onc(-c2c(Cl)cccc2Cl)c1C(=O)N[C@@H]1C(=O)N2[... dicloxacillin penicillins J01CF - Beta-lactamase resistant penicillins Cc1onc(-c2c(Cl)cccc2Cl)c1C(=O)N[C@@H]1C(=O)N2[... [C][C][O][N][=C][Branch1][=N][C][=C][Branch1][... InChI=1S/C19H17Cl2N3O5S/c1-7-10(12(23-29-7)11-... YFAGHNZHGGCZAX-JKIFEVAISA-N 
# Specifying only the standard SMILES column
df_full = full_data["standard_smiles"]
df_full
0     CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
1     CON(C)Cc1ccc(O)c2c1C[C@H]1C[C@H]3[C@H](N(C)C)C...
2     COc1c(N2CCNC(C)C2)c(F)cc2c(=O)c(C(=O)O)cn(C3CC...
3     CNC(C)[C@@H]1CC[C@@H](N)[C@@H](O[C@H]2[C@H](O)...
4     Cc1onc(-c2c(Cl)cccc2Cl)c1C(=O)N[C@@H]1C(=O)N2[...
                            ...                        
70                                Nc1ccc(S(N)(=O)=O)cc1
71     CC1COc2c(N3CCN(C)CC3)c(F)cc3c(=O)c(C(=O)O)cn1c23
72     O=C(/C=C/c1ccc(O)cc1)C1CCC/C(=C\c2ccc(O)cc2)C1=O
73            O=C(CCc1ccc(O)cc1)c1cccc(Cc2ccc(O)cc2)c1O
74           O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O
Name: standard_smiles, Length: 75, dtype: object
# Convert the standard SMILES into RDKit molecules
mol_full = [dm.to_mol(smile) for smile in df_full]


Aligning all the scaffolds

Here, all the scaffolds from both dataframes of compounds were aligned by using Datamol’s auto_align_many(). The images of all the aligned molecules were generated in the end. The compound structures did re-align, but unfortunately it only showed up to a maximum of 50 compounds only (the maximum number of molecules to be shown in Datamol was set to 32 as default; this number was pushed up to and truncated at 50 in the warning message from RDKit when attempting to run a total of 75 compounds using the Datamol library, without looking further into other ways to alter this for now).

aligned_list = dm.align.auto_align_many(mol_full, partition_method = "anon-scaffold")
dm.to_image(aligned_list, mol_size = (400, 400), max_mols = 50)

An attempt to combine Datamol’s auto_align_many() with mols2grid library was shown below. Unfortunately, the compounds did not re-align but all 75 compounds were shown in the grids.

mols2grid.display(aligned_list)
mols2grid library

Since I’ve started using mols2grid here, I thought to show an example of this library by combining all 75 compounds using the pre-processed standard SMILES in the grids with corresponding compound names. The resulting table provided a clear overview of all the compounds, with useful options to select or filter compounds. Obviously, other molecular properties or experimental results could be added into the table for other uses.

# Full dataset of 75 compounds 
mols2grid.display(full_data, smiles_col = "standard_smiles", subset = ["img", "mols2grid-id", "names"])


scaffold_finder library

Rather than only trying out Datamol only, I also thought to try out the scaffold_finder library after reading this Jupyter notebook by Patrick Walters. The GitHub repository of his other useful cheminformatics tutorials can be found here. His blog is here. Without surprises, this post was also inspired by this Jupyter notebook on “Identifying scaffolds in a set of molecules”, with some hope to expand on it a bit more.

Below were my notes on how to use this particular library.

Step 1: Add “mol” column to full_data dataframe (this was needed in order to use the functions from scaffold_finder library, which was also built based on RDKit)

full_data["mol"] = full_data.standard_smiles.apply(Chem.MolFromSmiles)

Step 2: Change column names of “standard_smiles” to “SMILES” & “names” to “Name” to match with scaffold_finder library functions with set column names (or other way round, by changing the names in the function library)

# Note: New column name "SMILES" contains standardised SMILES (old column name as "standard_smiles")
full_data = full_data.rename(columns = {"standard_smiles": "SMILES", "names": "Name"})
full_data.head()
smiles Name USAN Definition Level 4 ATC Codes SMILES selfies inchi inchikey mol
0 CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N... cefepime cephalosporins J01DE - Fourth-generation cephalosporins CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N... [C][O][/N][=C][Branch2][Ring2][#Branch2][\C][=... InChI=1S/C19H24N6O5S2/c1-25(5-3-4-6-25)7-10-8-... HVFLCNVBZFFHBT-ZKDACBOMSA-N <rdkit.Chem.rdchem.Mol object at 0x134307680>
1 CON(C)Cc1ccc(O)c2c1C[C@H]1C[C@H]3[C@H](N(C)C)C... sarecycline antibiotics (tetracycline derivatives) J01AA - Tetracyclines CON(C)Cc1ccc(O)c2c1C[C@H]1C[C@H]3[C@H](N(C)C)C... [C][O][N][Branch1][C][C][C][C][=C][C][=C][Bran... InChI=1S/C24H29N3O8/c1-26(2)18-13-8-11-7-12-10... PQJQFLNBMSCUSH-SBAJWEJLSA-N <rdkit.Chem.rdchem.Mol object at 0x134306c00>
2 COc1c(N2CCNC(C)C2)c(F)cc2c(=O)c(C(=O)O)cn(C3CC... gatifloxacin antibacterials (quinolone derivatives) S01AE - Fluoroquinolones | J01MA - Fluoroquino... COc1c(N2CCNC(C)C2)c(F)cc2c(=O)c(C(=O)O)cn(C3CC... [C][O][C][=C][Branch1][N][N][C][C][N][C][Branc... InChI=1S/C19H22FN3O4/c1-10-8-22(6-5-21-10)16-1... XUBOMFCQGDBHNK-UHFFFAOYSA-N <rdkit.Chem.rdchem.Mol object at 0x133e729d0>
3 CNC(C)[C@@H]1CC[C@@H](N)[C@@H](O[C@H]2[C@H](O)... gentamicin antibiotics (Micromonospora strains) S01AA - Antibiotics | S02AA - Antiinfectives |... CNC(C)[C@@H]1CC[C@@H](N)[C@@H](O[C@H]2[C@H](O)... [C][N][C][Branch1][C][C][C@@H1][C][C][C@@H1][B... InChI=1S/C21H43N5O7.C20H41N5O7.C19H39N5O7/c1-9... NPEFREDMMVQEPL-RWPARATISA-N <rdkit.Chem.rdchem.Mol object at 0x133e72500>
4 Cc1onc(-c2c(Cl)cccc2Cl)c1C(=O)N[C@@H]1C(=O)N2[... dicloxacillin penicillins J01CF - Beta-lactamase resistant penicillins Cc1onc(-c2c(Cl)cccc2Cl)c1C(=O)N[C@@H]1C(=O)N2[... [C][C][O][N][=C][Branch1][=N][C][=C][Branch1][... InChI=1S/C19H17Cl2N3O5S/c1-7-10(12(23-29-7)11-... YFAGHNZHGGCZAX-JKIFEVAISA-N <rdkit.Chem.rdchem.Mol object at 0x133e719a0>

Step 3: Identify scaffolds

The find_scaffolds() function was kindly borrowed from the scaffold_finder library as mentioned above. The scaffold_finder_test.py was the modified version, as I’ve used a different dataset here.

mol_df, scaffold_df = find_scaffolds(full_data)

Below was a quick overview of the mol_df, showing scaffolds in SMILES, number of atoms, number of R groups, names of compounds and the standardised SMILES of the compounds.

mol_df
Scaffold NumAtoms NumRgroupgs Name SMILES
0 C[N+]1(CC2=C(C(=O)[O-])N3C(=O)[C@@H](NC(=O)/C(... 31 1 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
1 C[N+]1(CC2=CN3C(=O)[C@@H](NC(=O)C(=NO)c4csc(N)... 28 2 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
2 Nc1nc(C(=NO)C(=O)N[C@@H]2C(=O)N3C=CCS[C@H]23)cs1 21 3 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
3 CC1=CN2C(=O)[C@@H](NC(=O)C(=NO)c3csc(N)n3)[C@H... 22 3 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
4 Nc1nc(C(=NO)C(=O)N[C@@H]2C(=O)N3C=C(C[NH+]4CCC... 27 3 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
... ... ... ... ... ...
15 O=C(CCc1ccccc1)C1CCCC(Cc2ccccc2)C1=O 24 2 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O
16 O=C1CCCCC1C(=O)CCc1ccc(O)cc1 18 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O
19 CC1CCCC(C(=O)CCc2ccc(O)cc2)C1=O 19 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O
21 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccccc2)C1=O 25 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O
22 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O 26 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O

5320 rows × 5 columns

Again to have a quick look at the scaffolds of all 75 compounds, along with counts of each scaffold and number of atoms in each scaffold.

scaffold_df
Scaffold Count NumAtoms
1156 CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@H]2C(O)=C3C(... 7 29
1101 CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C... 6 30
173 CC(=O)N[C@@H]1C(=O)N2[C@@H]1SC(C)(C)[C@@H]2C(=O)O 5 17
174 CC(=O)N[C@@H]1C(=O)N2[C@@H]1SC(C)[C@@H]2C(=O)O 5 16
552 CC1(C)S[C@@H]2[C@H](NC=O)C(=O)N2[C@H]1C(=O)O 5 16
... ... ... ...
3684 Cc1nccn1C 1 7
4456 Nc1ccccc1 1 7
4775 O=P(O)(O)[C@@H]1CO1 1 7
3722 Cn1ccnc1 1 6
4807 c1ccccc1 1 6

4808 rows × 3 columns

Step 4: Display all scaffolds in mols2grid, which helped to identify the scaffold with the highest frequency (counts) of occurence in the dataset.

mols2grid.display(scaffold_df, smiles_col = "Scaffold", subset = ["img", "Count"])


Test datasets

These were my sample datasets for later use in the section on “Reading and querying multiple scaffolds in SMILES strings”.

Below was the first test dataset on the top 2 scaffolds with highest frequency of appearance in the full dataframe.

# Scaffold of anti-infective with highest count
count_top1_scaffold = scaffold_df.Scaffold.values[0]
# Scaffold of anti-infective with the second highest count
count_top2_scaffold = scaffold_df.Scaffold.values[1]
# Combine above scaffolds into a list
count_top_scaffold = list((count_top1_scaffold, count_top2_scaffold))
count_top_scaffold
['CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@H]2C(O)=C3C(=O)c4c(O)cccc4C[C@H]3C[C@H]21',
 'CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C3C(=O)c4c(O)cccc4C[C@H]3C[C@@H]12']

Then this was the second sample dataset, purely on the scaffolds for the antibiotic called, “cefepime”.

cefe_scaffolds = pl.from_pandas(mol_df).filter(pl.col("Name") == "cefepime")
cefe_scaffolds.head()
shape: (5, 5)
Scaffold NumAtoms NumRgroupgs Name SMILES
str i64 i64 str str
"C[N+]1(CC2=C(C... 31 1 "cefepime" "CO/N=C(\C(=O)N...
"C[N+]1(CC2=CN3... 28 2 "cefepime" "CO/N=C(\C(=O)N...
"Nc1nc(C(=NO)C(... 21 3 "cefepime" "CO/N=C(\C(=O)N...
"CC1=CN2C(=O)[C... 22 3 "cefepime" "CO/N=C(\C(=O)N...
"Nc1nc(C(=NO)C(... 27 3 "cefepime" "CO/N=C(\C(=O)N...


Focus on FtsZ compound 3

Compound 3 was the compound found in the paper to have targeted the FtsZ proteins in Gram positive pathogens such as Streptococcus pneumoniae with more pronounced activities than its predecessor e.g. compound 1. So this section aimed to look into all of compound 3’s scaffolds.

# For ease of dataframe manipulation, decided to convert Pandas df into a Polars one (just my personal preference as I've used Polars more lately)
# then filtered out all the scaffolds for compound 3 & saved it as an independent dataframe
cpd3_scaffolds = pl.from_pandas(mol_df).filter(pl.col("Name") == "Compound_3")
cpd3_scaffolds
shape: (14, 5)
Scaffold NumAtoms NumRgroupgs Name SMILES
str i64 i64 str str
"O=CC1CCCC(Cc2c... 17 1 "Compound_3" "O=C(CCc1ccc(O)...
"O=CC1CCCC(Cc2c... 16 2 "Compound_3" "O=C(CCc1ccc(O)...
"CC(=O)C1CCCC(C... 18 1 "Compound_3" "O=C(CCc1ccc(O)...
"CC(=O)C1CCCC(C... 17 2 "Compound_3" "O=C(CCc1ccc(O)...
"CCC(=O)C1CCCC(... 19 1 "Compound_3" "O=C(CCc1ccc(O)...
"CCC(=O)C1CCCC(... 18 2 "Compound_3" "O=C(CCc1ccc(O)...
"O=C(CCc1ccccc1... 25 1 "Compound_3" "O=C(CCc1ccc(O)...
"O=C1CCCCC1C(=O... 17 2 "Compound_3" "O=C(CCc1ccc(O)...
"CC1CCCC(C(=O)C... 18 2 "Compound_3" "O=C(CCc1ccc(O)...
"O=C(CCc1ccccc1... 24 2 "Compound_3" "O=C(CCc1ccc(O)...
"O=C1CCCCC1C(=O... 18 1 "Compound_3" "O=C(CCc1ccc(O)...
"CC1CCCC(C(=O)C... 19 1 "Compound_3" "O=C(CCc1ccc(O)...
"O=C(CCc1ccc(O)... 25 1 "Compound_3" "O=C(CCc1ccc(O)...
"O=C(CCc1ccc(O)... 26 1 "Compound_3" "O=C(CCc1ccc(O)... 
# Convert Polars df into a Pandas one 
# and use mols2grid to show the 2D images of compound 3 scaffolds
# Total of 14 different scaffolds
cpd3_scaffolds = cpd3_scaffolds.to_pandas()
mols2grid.display(cpd3_scaffolds, smiles_col = "Scaffold")
Testing compound 3 scaffolds using scaffold_finder library

At this stage, I sort of had an idea of wanting to compare all 14 compound 3 scaffolds against all 75 molecules including ChEMBL-curated prescription-only anti-bacterials.

I tried the get_molecules_with_scaffold() function from scaffold_finder library but didn’t exactly get what I hoped to achieve. I played around a bit and noticed it was really designed for spotting a single target scaffold with highest counts in the data set. I was hoping to parse multiple scaffolds actually, or imagining there might be situations where we might want to do this.

I started trialling with one scaffold anyway as shown below on the get_molecule_with_scaffold() function from the scaffold_finder library.

# Trial single scaffold first
scaffold_test = cpd3_scaffolds.Scaffold.values[0]
scaffold_test
'O=CC1CCCC(Cc2ccc(O)cc2)C1=O'
s_smiles_list, chem_mol_df = get_molecules_with_scaffold(scaffold_test, mol_df, full_data)
s_smiles_list
array(['O=C1C(Cc2ccc(O)cc2)CCCC1C(=O)[*:1]'], dtype=object)
# Showing only compound 3 as a distinctive compound (no other molecules with similar scaffold)
chem_mol_df
SMILES Name
0 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O Compound_3


Reading and querying multiple scaffolds in SMILES strings

I also tried to tweak the get_molecule_with_scaffolds() function but realised it might be even better to write my own function to tailor to my need. Therefore, I wrote a small and simple function which would read and query multiple scaffolds of small molecules in SMILES string formats against a dataframe (with information showing scaffolds in SMILES, number of atoms, number of R groups, names of compounds and also the SMILES of the compounds).

At first, I started with reading all 14 scaffolds of compound 3 by using the values index method on cpd3_scaffolds dataframe that included all 14 scaffolds of compound 3.

# Trial feeding all 14 SMILES
scaffold_cpd3_all = cpd3_scaffolds.Scaffold.values[:]
scaffold_cpd3_all
array(['O=CC1CCCC(Cc2ccc(O)cc2)C1=O', 'O=CC1CCCC(Cc2ccccc2)C1=O',
       'CC(=O)C1CCCC(Cc2ccc(O)cc2)C1=O', 'CC(=O)C1CCCC(Cc2ccccc2)C1=O',
       'CCC(=O)C1CCCC(Cc2ccc(O)cc2)C1=O', 'CCC(=O)C1CCCC(Cc2ccccc2)C1=O',
       'O=C(CCc1ccccc1)C1CCCC(Cc2ccc(O)cc2)C1=O',
       'O=C1CCCCC1C(=O)CCc1ccccc1', 'CC1CCCC(C(=O)CCc2ccccc2)C1=O',
       'O=C(CCc1ccccc1)C1CCCC(Cc2ccccc2)C1=O',
       'O=C1CCCCC1C(=O)CCc1ccc(O)cc1', 'CC1CCCC(C(=O)CCc2ccc(O)cc2)C1=O',
       'O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccccc2)C1=O',
       'O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O'], dtype=object)

Then I thought about how every time if we’d want to convert any molecules from SMILES to RDKit molecules, we really had to have a “mol” column set up, so that was what I did below.

cpd3_scaffolds["mol"] = cpd3_scaffolds.Scaffold.apply(Chem.MolFromSmiles)

The dataframe would look like this now.

cpd3_scaffolds
Scaffold NumAtoms NumRgroupgs Name SMILES mol
0 O=CC1CCCC(Cc2ccc(O)cc2)C1=O 17 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O <rdkit.Chem.rdchem.Mol object at 0x13580d3f0>
1 O=CC1CCCC(Cc2ccccc2)C1=O 16 2 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O <rdkit.Chem.rdchem.Mol object at 0x13580d4d0>
2 CC(=O)C1CCCC(Cc2ccc(O)cc2)C1=O 18 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O <rdkit.Chem.rdchem.Mol object at 0x13580dd20>
3 CC(=O)C1CCCC(Cc2ccccc2)C1=O 17 2 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O <rdkit.Chem.rdchem.Mol object at 0x13580e880>
4 CCC(=O)C1CCCC(Cc2ccc(O)cc2)C1=O 19 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O <rdkit.Chem.rdchem.Mol object at 0x13580d930>
5 CCC(=O)C1CCCC(Cc2ccccc2)C1=O 18 2 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O <rdkit.Chem.rdchem.Mol object at 0x13580ea40>
6 O=C(CCc1ccccc1)C1CCCC(Cc2ccc(O)cc2)C1=O 25 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O <rdkit.Chem.rdchem.Mol object at 0x13580ca50>
7 O=C1CCCCC1C(=O)CCc1ccccc1 17 2 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O <rdkit.Chem.rdchem.Mol object at 0x13580fbc0>
8 CC1CCCC(C(=O)CCc2ccccc2)C1=O 18 2 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O <rdkit.Chem.rdchem.Mol object at 0x13580f4c0>
9 O=C(CCc1ccccc1)C1CCCC(Cc2ccccc2)C1=O 24 2 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O <rdkit.Chem.rdchem.Mol object at 0x13580f060>
10 O=C1CCCCC1C(=O)CCc1ccc(O)cc1 18 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O <rdkit.Chem.rdchem.Mol object at 0x13580c200>
11 CC1CCCC(C(=O)CCc2ccc(O)cc2)C1=O 19 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O <rdkit.Chem.rdchem.Mol object at 0x13580d770>
12 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccccc2)C1=O 25 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O <rdkit.Chem.rdchem.Mol object at 0x13580de70>
13 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O 26 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O <rdkit.Chem.rdchem.Mol object at 0x13580e2d0>

Then perhaps I would place all of these compound 3 scaffolds into an object.

cpd3_mols = cpd3_scaffolds["mol"]
cpd3_mols
0     <rdkit.Chem.rdchem.Mol object at 0x13580d3f0>
1     <rdkit.Chem.rdchem.Mol object at 0x13580d4d0>
2     <rdkit.Chem.rdchem.Mol object at 0x13580dd20>
3     <rdkit.Chem.rdchem.Mol object at 0x13580e880>
4     <rdkit.Chem.rdchem.Mol object at 0x13580d930>
5     <rdkit.Chem.rdchem.Mol object at 0x13580ea40>
6     <rdkit.Chem.rdchem.Mol object at 0x13580ca50>
7     <rdkit.Chem.rdchem.Mol object at 0x13580fbc0>
8     <rdkit.Chem.rdchem.Mol object at 0x13580f4c0>
9     <rdkit.Chem.rdchem.Mol object at 0x13580f060>
10    <rdkit.Chem.rdchem.Mol object at 0x13580c200>
11    <rdkit.Chem.rdchem.Mol object at 0x13580d770>
12    <rdkit.Chem.rdchem.Mol object at 0x13580de70>
13    <rdkit.Chem.rdchem.Mol object at 0x13580e2d0>
Name: mol, dtype: object

At this stage, nothing really clicked at the moment, but then I thought about how Datamol was built on top of RDKit and also how a few other cheminformatics posts I’ve read before utilised the functions in RDKit, so it was time to look deeper in RDKit to search for methods with the intended purpose in mind. I then found the SmilesWriter() method from RDKit after reading a few online references.

I’ve found out that:

  • To write multiple SMILES into a .smi file, use SmilesWriter()

  • To read a set of SMILES from a .smi file, use SmilesMolSupplier()

Acknowledgement of a useful link I’ve found online which had helped me to figure out how to save multiple SMILES strings in a .smi file.

# Figured out how to save multiple SMILES as a text file 
cpd3 = SmilesWriter('cpd3.smi')

# Note: saving multiple SMILES strings from RDKit mol objects (cpd3_mols)
for s in cpd3_mols:
  cpd3.write(s)
cpd3.close()


Function for saving multiple SMILES strings as a .smi file

So based on the ideas in the previous section, I came up with the following simple function to save multiple SMILES strings as a .smi file.

def save_smiles_strings(df, file_name):

  # Create a RDKit mol column in the dataframe
  df["mol"] = df.Scaffold.apply(Chem.MolFromSmiles)

  # Save the "mol" column with target scaffolds as an object
  smiles_mols = df["mol"]

  # Use RDKit's SmilesWriter() to write the smiles strings from mol objects
  # Specify file name for .smi file, which will be stored in the working directory
  smiles = SmilesWriter(f"{file_name}.smi")
  
  # Use a for loop to iterate through each SMILES string in the dataframe
  for s in smiles_mols:
    smiles.write(s)
  smiles.close()


Testing on the function

Here I used one of the dataframes I’ve saved earlier, cefe_scaffolds, to test this function on saving multiple SMILES into a file. Since cefe_scaffolds dataframe was a Polars dataframe from earlier, it needed to be converted into a Pandas dataframe in order to be compatible with RDKit, which was used in the function.

# Convert Polars dataframe into a Pandas one
cefe_scaffolds = cefe_scaffolds.to_pandas()

# Running function on cefe_scaffolds dataframe
# First parameter - dataframe to be used
# Second parameter - file name for the SMILES strings saved
save_smiles_strings(cefe_scaffolds, "cefe")

A .smi file with the name “cefe.smi” should appear in the working directory after running the function.

Now, the next stage would be to parse these SMILES strings and save them as a list. I actually worked backwards here where I looked into the Pandas.query() method first, and looked into options I could have on reading and checking for matches in multiple strings. To be able to read multiple strings in one go, a list would be suitable to carry out the matching queries (note: in the scaffold_finder library, this dataframe query method was also used in its “find_scaffolds” and “get_molecules_with_scaffold” functions).

An example of Pandas.query() tests:

# Using the test dataset from earlier - list of top two scaffolds with highest frequency of occurrences from ChEMBL dataset
count_top_scaffold
['CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@H]2C(O)=C3C(=O)c4c(O)cccc4C[C@H]3C[C@H]21',
 'CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C3C(=O)c4c(O)cccc4C[C@H]3C[C@@H]12']
# To demonstrate that querying the two top scaffolds 
# will bring back all the anti-bacterials with the same scaffold
match_df = mol_df.query("Scaffold in @count_top_scaffold")
match_df
Scaffold NumAtoms NumRgroupgs Name SMILES
102 CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C... 30 1 sarecycline CON(C)Cc1ccc(O)c2c1C[C@H]1C[C@H]3[C@H](N(C)C)C...
128 CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@H]2C(O)=C3C(... 29 2 sarecycline CON(C)Cc1ccc(O)c2c1C[C@H]1C[C@H]3[C@H](N(C)C)C...
92 CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@H]2C(O)=C3C(... 29 3 tigecycline CN(C)c1cc(NC(=O)CNC(C)(C)C)c(O)c2c1C[C@H]1C[C@...
115 CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@H]2C(O)=C3C(... 29 3 demeclocycline CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C...
129 CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C... 30 2 demeclocycline CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C...
39 CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C... 30 1 minocycline CN(C)c1ccc(O)c2c1C[C@H]1C[C@H]3[C@H](N(C)C)C(O...
65 CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@H]2C(O)=C3C(... 29 2 minocycline CN(C)c1ccc(O)c2c1C[C@H]1C[C@H]3[C@H](N(C)C)C(O...
115 CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@H]2C(O)=C3C(... 29 3 tetracycline CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C...
129 CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C... 30 2 tetracycline CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C...
171 CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@H]2C(O)=C3C(... 29 3 eravacycline CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C...
201 CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C... 30 2 eravacycline CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C...
172 CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C... 30 3 oxytetracycline CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C...
92 CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@H]2C(O)=C3C(... 29 3 omadacycline CN(C)c1cc(CNCC(C)(C)C)c(O)c2c1C[C@H]1C[C@H]3[C...

Here, all tetracycline antibiotics were brought up in the resultant dataframe.

As an aside from what I wanted to do, I also learnt a small trick about how to get the number of atoms in the file with multiple SMILES strings.

# Sample use of SmilesMolSupplier & GetNumAtoms()
suppl = SmilesMolSupplier('cpd3.smi')

nMols = len(suppl)

for i in range(nMols):

  a = suppl[i].GetNumAtoms()
  print(a)
17
16
18
17
19
18
25
17
18
24
18
19
25
26

Now, back to where I was meant to continue working, I wanted to convert these SMILES strings into RDKit molecules first.

# Reading cpd3.smi SMILES strings in text file as RDKit mol objects
suppl = SmilesMolSupplier("cpd3.smi")
suppl
<rdkit.Chem.rdmolfiles.SmilesMolSupplier at 0x134bf5c10>

This was followed by converting the “mol” objects into SMILES strings, so that we could save each SMILES string into a list.

# Initialise an empty list
list = []

for mol in suppl:
   # Convert RDKit mol objects into SMILES strings
   m = Chem.MolToSmiles(mol)
   # Add each SMILES read from filename.smi into the empty list
   list.append(m)

list
['O=CC1CCCC(Cc2ccc(O)cc2)C1=O',
 'O=CC1CCCC(Cc2ccccc2)C1=O',
 'CC(=O)C1CCCC(Cc2ccc(O)cc2)C1=O',
 'CC(=O)C1CCCC(Cc2ccccc2)C1=O',
 'CCC(=O)C1CCCC(Cc2ccc(O)cc2)C1=O',
 'CCC(=O)C1CCCC(Cc2ccccc2)C1=O',
 'O=C(CCc1ccccc1)C1CCCC(Cc2ccc(O)cc2)C1=O',
 'O=C1CCCCC1C(=O)CCc1ccccc1',
 'CC1CCCC(C(=O)CCc2ccccc2)C1=O',
 'O=C(CCc1ccccc1)C1CCCC(Cc2ccccc2)C1=O',
 'O=C1CCCCC1C(=O)CCc1ccc(O)cc1',
 'CC1CCCC(C(=O)CCc2ccc(O)cc2)C1=O',
 'O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccccc2)C1=O',
 'O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O']


Function for converting .smi file into a list to query and match scaffolds of interests

I then came up with the next function that would feed multiple scaffolds into a Pandas.query() to match the strings, meaning we could compare the scaffolds with each other in a dataframe.

def query_scaffolds_via_smiles(filename):

    # Initialise an empty list
    list = []
    # Use SmilesMolSupplier() from RDKit to read in the SMILES strings stored in .smi file
    suppl = SmilesMolSupplier(filename)
    # Use a for loop to iterate through the SMILES strings
    for mol in suppl:
        # Convert RDKit mol objects into SMILES strings
        m = Chem.MolToSmiles(mol)
        # Add each SMILES read from filename.smi into the empty list
        list.append(m)
        # Compare the SMILES with the scaffold column in dataframe
        scaffold_match_df = mol_df.query("Scaffold in @list")
        
    return scaffold_match_df


Testing on the function

Below was a test for this query_scaffolds_via_smiles() function using the previously made “cpd3.smi” file.

To show that compound 3 scaffolds literally only existed in compound 3 and not in any other prescription-only anti-bacterials based on the ChEMBL-extracted anti-infective dataset only (note: other sources not checked at this stage).

# Testing query_scaffolds_via_smiles() function
query_scaffolds_via_smiles("cpd3.smi")
Scaffold NumAtoms NumRgroupgs Name SMILES
0 O=CC1CCCC(Cc2ccc(O)cc2)C1=O 17 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O
4 O=CC1CCCC(Cc2ccccc2)C1=O 16 2 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O
5 CC(=O)C1CCCC(Cc2ccc(O)cc2)C1=O 18 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O
8 CC(=O)C1CCCC(Cc2ccccc2)C1=O 17 2 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O
9 CCC(=O)C1CCCC(Cc2ccc(O)cc2)C1=O 19 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O
11 CCC(=O)C1CCCC(Cc2ccccc2)C1=O 18 2 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O
12 O=C(CCc1ccccc1)C1CCCC(Cc2ccc(O)cc2)C1=O 25 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O
13 O=C1CCCCC1C(=O)CCc1ccccc1 17 2 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O
14 CC1CCCC(C(=O)CCc2ccccc2)C1=O 18 2 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O
15 O=C(CCc1ccccc1)C1CCCC(Cc2ccccc2)C1=O 24 2 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O
16 O=C1CCCCC1C(=O)CCc1ccc(O)cc1 18 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O
19 CC1CCCC(C(=O)CCc2ccc(O)cc2)C1=O 19 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O
21 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccccc2)C1=O 25 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O
22 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O 26 1 Compound_3 O=C(CCc1ccc(O)cc1)C1CCCC(Cc2ccc(O)cc2)C1=O

Then I also tested on “cefe.smi” created before.

# Test on cefe.smi
query_scaffolds_via_smiles("cefe.smi")
Scaffold NumAtoms NumRgroupgs Name SMILES
0 C[N+]1(CC2=C(C(=O)[O-])N3C(=O)[C@@H](NC(=O)/C(... 31 1 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
1 C[N+]1(CC2=CN3C(=O)[C@@H](NC(=O)C(=NO)c4csc(N)... 28 2 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
2 Nc1nc(C(=NO)C(=O)N[C@@H]2C(=O)N3C=CCS[C@H]23)cs1 21 3 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
3 CC1=CN2C(=O)[C@@H](NC(=O)C(=NO)c3csc(N)n3)[C@H... 22 3 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
4 Nc1nc(C(=NO)C(=O)N[C@@H]2C(=O)N3C=C(C[NH+]4CCC... 27 3 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
5 C[N+]1(CC2=CN3C(=O)[C@@H](NC(=O)C=NO)[C@H]3SC2... 22 3 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
6 C[N+]1(CC2=CN3C(=O)[C@@H](NC(=O)C(=NO)c4cscn4)... 27 3 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
7 Nc1nc(C(=NO)C(=O)N[C@@H]2C(=O)N3C(C(=O)[O-])=C... 24 2 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
8 O=C([O-])C1=CCS[C@@H]2[C@H](NC(=O)C(=NO)c3cscn... 23 3 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
9 CC1=C(C(=O)[O-])N2C(=O)[C@@H](NC(=O)C(=NO)c3cs... 25 2 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
10 CC1=C(C(=O)[O-])N2C(=O)[C@@H](NC(=O)C=NO)[C@H]... 19 3 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
11 CC1=C(C(=O)[O-])N2C(=O)[C@@H](NC(=O)C(=NO)c3cs... 24 3 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
12 Nc1nc(C(=NO)C(=O)N[C@@H]2C(=O)N3C(C(=O)[O-])=C... 30 2 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
13 O=C(C=NO)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[NH+]3... 24 3 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
14 O=C([O-])C1=C(C[NH+]2CCCC2)CS[C@@H]2[C@H](NC(=... 29 3 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
15 C[N+]1(CC2=C(C(=O)[O-])N3C(=O)[C@@H](NC(=O)C=N... 25 2 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
16 C[N+]1(CC2=C(C(=O)[O-])N3C(=O)[C@@H](NC(=O)C(=... 30 2 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
17 CO/N=C(\C(=O)N[C@@H]1C(=O)N2C=C(C[N+]3(C)CCCC3... 29 1 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
18 CON=C(C(=O)N[C@@H]1C(=O)N2C=CCS[C@H]12)c1csc(N)n1 22 2 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
19 CON=C(C(=O)N[C@@H]1C(=O)N2C=CCS[C@H]12)c1cscn1 21 3 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
20 CON=C(C(=O)N[C@@H]1C(=O)N2C=C(C)CS[C@H]12)c1cs... 23 2 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
21 CON=C(C(=O)N[C@@H]1C(=O)N2C=C(C)CS[C@H]12)c1cscn1 22 3 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
22 CON=C(C(=O)N[C@@H]1C(=O)N2C=C(C[NH+]3CCCC3)CS[... 28 2 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
23 CON=CC(=O)N[C@@H]1C(=O)N2C=C(C[NH+]3CCCC3)CS[C... 22 3 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
24 CON=C(C(=O)N[C@@H]1C(=O)N2C=C(C[NH+]3CCCC3)CS[... 27 3 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
25 CON=CC(=O)N[C@@H]1C(=O)N2C=C(C[N+]3(C)CCCC3)CS... 23 2 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
26 CON=C(C(=O)N[C@@H]1C(=O)N2C=C(C[N+]3(C)CCCC3)C... 28 2 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
27 CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=CCS[C... 25 1 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
28 CON=C(C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=CCS[C@H... 25 1 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
30 CON=C(C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=CCS[C@H... 24 2 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
31 CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C)C... 26 1 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
32 CON=C(C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C)CS[... 26 1 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
33 CON=CC(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C)CS[C... 20 2 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
34 CON=C(C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C)CS[... 25 2 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
35 CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N... 31 1 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
36 CON=CC(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[NH+]... 25 2 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
37 CON=C(C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[NH+... 30 2 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
38 CO/N=C\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N+... 26 1 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
39 CON=CC(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N+]3... 26 1 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
40 CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N... 31 1 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
41 CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N... 32 1 cefepime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=C(C[N...
2 Nc1nc(C(=NO)C(=O)N[C@@H]2C(=O)N3C=CCS[C@H]23)cs1 21 3 cefixime C=CC1=C(C(=O)O)N2C(=O)[C@@H](NC(=O)/C(=N\OCC(=...
3 CON=C(C(=O)N[C@@H]1C(=O)N2C=CCS[C@H]12)c1csc(N)n1 22 3 cefixime C=CC1=C(C(=O)O)N2C(=O)[C@@H](NC(=O)/C(=N\OCC(=...
2 CC1=CN2C(=O)[C@@H](NC(=O)C(=NO)c3csc(N)n3)[C@H... 22 3 ceftriaxone CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)O)=C(CSc3nc...
24 CON=C(C(=O)N[C@@H]1C(=O)N2C=C(C)CS[C@H]12)c1cs... 23 2 ceftriaxone CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)O)=C(CSc3nc...
25 CON=C(C(=O)N[C@@H]1C(=O)N2C=C(C)CS[C@H]12)c1cscn1 22 3 ceftriaxone CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)O)=C(CSc3nc...
1 Nc1nc(C(=NO)C(=O)N[C@@H]2C(=O)N3C=CCS[C@H]23)cs1 21 2 cefdinir C=CC1=C(C(=O)O)N2C(=O)[C@@H](NC(=O)/C(=N\O)c3c...
2 Nc1nc(C(=NO)C(=O)N[C@@H]2C(=O)N3C=CCS[C@H]23)cs1 21 3 cefotaxime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)O)=C(COC(C)...
3 CC1=CN2C(=O)[C@@H](NC(=O)C(=NO)c3csc(N)n3)[C@H... 22 3 cefotaxime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)O)=C(COC(C)...
19 CON=C(C(=O)N[C@@H]1C(=O)N2C=CCS[C@H]12)c1csc(N)n1 22 2 cefotaxime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)O)=C(COC(C)...
20 CON=C(C(=O)N[C@@H]1C(=O)N2C=CCS[C@H]12)c1cscn1 21 3 cefotaxime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)O)=C(COC(C)...
21 CON=C(C(=O)N[C@@H]1C(=O)N2C=C(C)CS[C@H]12)c1cs... 23 2 cefotaxime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)O)=C(COC(C)...
22 CON=C(C(=O)N[C@@H]1C(=O)N2C=C(C)CS[C@H]12)c1cscn1 22 3 cefotaxime CO/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)O)=C(COC(C)...
33 CC1=CN2C(=O)[C@@H](NC(=O)C(=NO)c3csc(N)n3)[C@H... 22 3 ceftazidime CC(C)(O/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=...
36 Nc1nc(C(=NO)C(=O)N[C@@H]2C(=O)N3C(C(=O)[O-])=C... 24 2 ceftazidime CC(C)(O/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=...
37 O=C([O-])C1=CCS[C@@H]2[C@H](NC(=O)C(=NO)c3cscn... 23 3 ceftazidime CC(C)(O/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=...
38 CC1=C(C(=O)[O-])N2C(=O)[C@@H](NC(=O)C(=NO)c3cs... 25 2 ceftazidime CC(C)(O/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=...
39 CC1=C(C(=O)[O-])N2C(=O)[C@@H](NC(=O)C(=NO)c3cs... 24 3 ceftazidime CC(C)(O/N=C(\C(=O)N[C@@H]1C(=O)N2C(C(=O)[O-])=...

Only cephalosporins were brought up as the result from this scaffold query, which matched with the antibiotic class of cefepime.


Some answers

There were no other prescription-only anti-bacterials from ChEMBL database with the same scaffold as compound 3 after comparing the SMILES strings between these selected scaffolds only. This was only limited to the dataset obtained from ChEMBL at this stage, with compound indications limited to anti-infectives for now. This might imply that the scaffold of compound 3 could be considered novel when comparing with molecules with similar indications. Obviously, this was too preliminary to confirm anything substantial, since there were no in vivo tests on finding out the efficacy, safety and toxicity of compound 3 apart from the in vitro experimental results mentioned in the paper. It probably would provide some ideas for scaffold hopping in hit compounds, or functional group (R-group) comparisons when looking for new compounds for synthesis. Overall, it was interesting to have a re-visit on this work using a more cheminformatics approach.


Afterthoughts

What I wanted to achieve in this post were:

  • To familiarise myself with Datamol and scaffold_finder libraries

  • To use Polars dataframe library for initial data wrangling along with Datamol Python library, and the later trial of scaffold_finder library. Using Polars would be a small degree only, as Datamol was likely written with Pandas in mind mostly (based on RDKit), while Pandas was also the more commonly used dataframe libray in many cheminformatics packages. Some people might prefer to stick with Pandas all the way, which I agree, but I’d just wanted to use Polars for the initial data wrangling only as I’ve been using it more lately

  • To reveal my thought process on building simple cheminformatics-related functions (this was unplanned, but kind of evolved while working on this post)

  • To show some evidence of my own growth from computational and medicinal chemistry with no code, to using data science tools with Python code to help guiding drug discovery projects

  • To mention that experimental validations will always be crucial for computational predictions, and since I had some experimental results from the paper, I thought to accompany it with some computational findings here

I hope I’ve at least achieved some these points in this post if not all.

Thanks for reading and looking forward to comments if any.

Back to top

Footnotes

  1. Simplified Molecular Input Line Entry Systems↩︎

  2. SELF-referencIng Embedded Strings↩︎

  3. International Chemical Identifier↩︎

]]> RDKit Datamol Python Pandas Polars Cheminformatics https://jhylin.github.io/Data_in_life_blog/posts/14_Scaffolds_in_small_molecules/chembl_anti-inf_data_prep_current.html Wed, 05 Jul 2023 12:00:00 GMT Shinylive app in Python Jennifer HY Lin https://jhylin.github.io/Data_in_life_blog/posts/13_Shiny_app_python/ShinyAppPy_PC_Cov19_app_embed_pyodide_http.html
Quick update

I’ve changed the way of importing a local text/csv file from manually copying-and-pasting to using pyodide.http.open_url() in the shinylive app, which works great and avoids the clumsy manual file input. I couldn’t quite grasp the code back then, but managed to get it this time when I re-visited the problem, while also figured out that I could use the raw content link for the file from my GitHub repository. This method was also inspired by the same user answering the query in this GitHub discussion. So I’ve basically trialled both ways as suggested, which have all worked.

Note: if importing binary files, use pyodide.http.pyfetch() instead - check out Pyodide for details and latest changes.


Shinylive app in action

Note: it may take a few minutes to load the app (code provided at the top, with app at the bottom).

#| '!! shinylive warning !!': |
#|   shinylive does not work in self-contained HTML documents.
#|   Please set `embed-resources: false` in your metadata.
#| standalone: true
#| components: [editor, viewer]
#| layout: vertical
#| viewerHeight: 420

## file: app.py
# ***Import all libraries or packages needed***
# Import shiny ui, app
from shiny import ui, App
# Import shinywidgets
from shinywidgets import output_widget, render_widget
# Import shinyswatch to add themes
#import shinyswatch
# Import plotly express
import plotly.express as px
# Import pandas
import pandas as pd
# Import pyodide http - for importing file via URL
import pyodide.http
from pyodide.http import open_url


# ***Specify data source***
# Using pyodide.http.open_url
df = pd.read_csv(open_url('https://raw.githubusercontent.com/jhylin/Data_in_life_blog/main/posts/13_Shiny_app_python/pc_cov_pd.csv'))


# User interface---
# Add inputs & outputs
app_ui = ui.page_fluid(
        # Add theme - seems to only work in VS code and shinyapps.io
        #shinyswatch.theme.superhero(),
        # Add heading
        ui.h3("Molecular properties of compounds used in COVID-19 clinical trials"),
        # Place selection boxes & texts in same row
        ui.row(
            # Divide the row into two columns
            # Column 1 - selection drop-down boxes x 2
            ui.column(
                4, ui.input_select(
                # Specify x variable input
                "x", label = "x axis:", 
                choices = ["Partition coefficients", 
                           "Complexity",
                           "Heavy atom count",
                           "Hydrogen bond donor count",
                           "Hydrogen bond acceptor count",
                           "Rotatable bond count",
                           "Molecular weight",
                           "Exact mass", 
                           "Polar surface area", 
                           "Total atom stereocenter count", 
                           "Total bond stereocenter count"],
                ), 
                ui.input_select(
                # Specify y variable input
                "y", label = "y axis:",
                choices = ["Partition coefficients", 
                           "Complexity",
                           "Heavy atom count",
                           "Hydrogen bond donor count",
                           "Hydrogen bond acceptor count",
                           "Rotatable bond count", 
                           "Molecular weight",
                           "Exact mass", 
                           "Polar surface area", 
                           "Total atom stereocenter count", 
                           "Total bond stereocenter count"]  
                )),
            # Column 2 - add texts regarding plots
            ui.column(
            8,
            ui.p("Select different molecular properties as x and y axes to produce a scatter plot."),
            ui.tags.ul(
                ui.tags.li(
                    """
                    Part_coef_group means groups of partition coefficient (xlogp) as shown in the legend on the right""" 
                ), 
                ui.tags.li(
                    """
                    Toggle each partition coefficient category by clicking on the group names"""
                ), 
                ui.tags.li(
                    """
                    Hover over each data point to see compound name and relevant molecular properties"""
                )
            )),
        # Output as a widget (interactive plot)
        output_widget("my_widget"), 
        # Add texts for data source
        ui.row(
            ui.p(
                """
                Data curated by PubChem, accessed from: https://pubchem.ncbi.nlm.nih.gov/#tab=compound&query=covid-19%20clinicaltrials (last access date: 30th Apr 2023)""" 
            )         
        ) 
    )
)


# Server---
# Add plotting code within my_widget function within the server function
def server(input, output, session):
    @output
    @render_widget
    def my_widget():
        fig = px.scatter(
            df, x = input.x(), y = input.y(),
            color = "Part_coef_group", 
            hover_name = "Compound name"
        )
        fig.layout.height = 400
        return fig
        
# Combine UI & server into Shiny app
app = App(app_ui, server)
Back to top ]]> Python Shiny Pandas Plotly PubChem Cheminformatics https://jhylin.github.io/Data_in_life_blog/posts/13_Shiny_app_python/ShinyAppPy_PC_Cov19_app_embed_pyodide_http.html Tue, 23 May 2023 12:00:00 GMT Shinylive app in Python Jennifer HY Lin https://jhylin.github.io/Data_in_life_blog/posts/13_Shiny_app_python/ShinyAppPy_PC_Cov19_app_embed.html
Coding for Shinylive app in Python

The entire code for this Shiny app in Python were written in VS Code initially. When I was looking for places to deploy the app, I then migrated the code to RStudio IDE (note: recently I saw Python code in Quarto documents being used in VS Code after I’ve deployed this app, so this might be another option).

For data preparation stage, please visit this post for details.


Trialling open-source large-language models

There have been a lot of hypes surrounding large-language models (LLMs) or generative pre-trained transformers (GPTs). I’m still somewhat both reserved and excited about them at present. The area I’m most concerned with is where are the relevant laws, regulations and ethics on using these tools in public domains, whether for private or commercial uses, and will they be country-specific or globally streamlined? While I have these questions in mind, I’m not denying the fact that they can be useful for a few well-known purposes in coding, such as requesting for regex code templates for text cleaning or asking for code snippets while programmers are busy with several tasks at the same time.

I thought to trial out the open-source LLMs a little bit when I was working on this app, since they were the open and transparent versions with potentials for personal or commercial uses. So I’ve tested about 3 different open-source LLMs, which were H2OGPT, HuggingChat and StableLM. I decided to only give minimal amount of prompts (I’ve given H2OGPT three prompts, and only one prompt each for either HuggingChat and StableLM), and the question was framed around providing a code outline for building a Shiny app using Python and Polars dataframe libary, and see what answers they could each provide.

All of them produced answers that were close to what I’ve asked for, i.e. I could see a code outline for a Shiny app.py file at first but it was more like R code, with one case mixing both Python and R in the code concurrently. However, none of them actually reached exactly how I’ve asked them to provide, perhaps if I’ve given more prompts then they might manage to provide answers closer to the request. Also, none of them were able to pick up the more recent Polars dataframe library (which were probably more prominent from 2021 onwards) and also Shiny in Python (which was also quite new as well, only out of alpha phase recently). All of them only showed code using Pandas dataframe libray, even though I did mention Polars in all of the question prompts. Once I get to use more of these open-source LLMs in the future, which currently are gaining more traction from what I’ve read online, I think I might try to write another post about them later.


Using Shiny in Python documentations

After trialling open-source LLMs a bit to help with the coding part (although not as helpful as I first imagined, but they still somewhat provided a rough code framework for Shiny apps), I completed the app.py script in the end by using mainly documents from Shiny for Python website. I still think thorough documentations for any products are still fundamental and inevitable, because these documentations will be sources to provide high-quality data for training these LLMs which means they’ll be of higher quality too.

Several links I’ve used and found to be very informative:


App deployment

After building the app in a workable condition, I started looking at where and how to deploy this Shiny app written in Python. Initially I used the easiest method as stated from the Shinylive link above, which was to deploy on Shinylive editor. One drawback for this was that the application URL was extremely long as the URL hash would store the entire app code inside. However, this was indeed one of the simplest ways to share the apps with others quickly, by simply providing the URL link with the intended audience.

I then went on to try another method where I spent one afternoon trying to figure out that I had to add an install_certificate.command for my operating system since I installed Python by using Homebrew in the first place (which some clever users might already know to avoid…), so that I could resolve the SSL certificate issue if I’m trying to connect with Shinyapps.io.

I also tried to follow another method from an earlier time which suggested to deploy the app directly on GitHub Pages, but since then the code was changed and updated to a different version and the old method no longer worked as nicely as mentioned (a lot of the code for Shiny in Python can still be experimental, and can change drastically, so check the sources to see the latest updates). I then stumbled upon another GitHub repository from the Quarto team at Posit/RStudio and looked into the possibility and tested it. Eventually, I settled on this method - embedding the app in a Quarto document, which provided both code and app itself on this very same webpage down below.

One of the downsides of the following code that I couldn’t quite get rid of yet was the poor file importation style. for loading a .csv file locally. I’ve added code annotations as comments to explain what I’ve done. The solution to import .csv file into app.py for the embedding version was from this discussion in the repository (thanks to that particular user who found this hacky way to do it). However, the plus side was that coding for the rest of the app was not bad at all and worked seamlessly. I’ll try to follow up later on whether importing .csv file for app embedding could be easier, perhaps this might be something the Quarto team is working right now.

The second downside was that shinyswatch package might not be functional yet for this embedding method (but most likely would work if deploying apps to shinyapps.io). This meant there would be no background visual themes yet for apps embedded in Quarto docs, but hopefully this would be possible in the future. Overall, I was amazed at how easy it was to build, deploy and embed a Shiny app in Python in a Quarto document.

Note: if using Shinylive editor, make sure to include a requirement.txt file if using extra Python packages such as Pandas or Plotly. For embedding apps in Quarto docs, this appears not to be compulsory and only optional when I tried it down below.

Code for this post (.qmd document) is available here.


Shinylive app in action

Note: it may take a few minutes to load the app (code provided at the top, with app at the bottom).

Update on 24/5/23: to avoid manual file input, see this updated post which uses pyodide.http.open_url() method (code change in editor section).

#| '!! shinylive warning !!': |
#|   shinylive does not work in self-contained HTML documents.
#|   Please set `embed-resources: false` in your metadata.
#| standalone: true
#| components: [editor, viewer]
#| layout: vertical
#| viewerHeight: 420

## file: app.py
# ***Import all libraries or packages needed***
# Import shiny ui, app
from shiny import ui, App
# Import shinywidgets
from shinywidgets import output_widget, render_widget
# Import shinyswatch to add themes
#import shinyswatch
# Import plotly express
import plotly.express as px
# Import pandas
import pandas as pd
from pathlib import Path


# User interface---
# Add inputs & outputs
app_ui = ui.page_fluid(
        # Add theme - seems to only work in VS code and shinyapps.io
        #shinyswatch.theme.superhero(),
        # Add heading
        ui.h3("Molecular properties of compounds used in COVID-19 clinical trials"),
        # Place selection boxes & texts in same row
        ui.row(
            # Divide the row into two columns
            # Column 1 - selection drop-down boxes x 2
            ui.column(
                4, ui.input_select(
                # Specify x variable input
                "x", label = "x axis:", 
                choices = ["Partition coefficients", 
                           "Complexity",
                           "Heavy atom count",
                           "Hydrogen bond donor count",
                           "Hydrogen bond acceptor count",
                           "Rotatable bond count",
                           "Molecular weight",
                           "Exact mass", 
                           "Polar surface area", 
                           "Total atom stereocenter count", 
                           "Total bond stereocenter count"],
                ), 
                ui.input_select(
                # Specify y variable input
                "y", label = "y axis:",
                choices = ["Partition coefficients", 
                           "Complexity",
                           "Heavy atom count",
                           "Hydrogen bond donor count",
                           "Hydrogen bond acceptor count",
                           "Rotatable bond count", 
                           "Molecular weight",
                           "Exact mass", 
                           "Polar surface area", 
                           "Total atom stereocenter count", 
                           "Total bond stereocenter count"]  
                )),
            # Column 2 - add texts regarding plots
            ui.column(
            8,
            ui.p("Select different molecular properties as x and y axes to produce a scatter plot."),
            ui.tags.ul(
                ui.tags.li(
                    """
                    Part_coef_group means groups of partition coefficient (xlogp) as shown in the legend on the right""" 
                ), 
                ui.tags.li(
                    """
                    Toggle each partition coefficient category by clicking on the group names"""
                ), 
                ui.tags.li(
                    """
                    Hover over each data point to see compound name and relevant molecular properties"""
                )
            )),
        # Output as a widget (interactive plot)
        output_widget("my_widget"), 
        # Add texts for data source
        ui.row(
            ui.p(
                """
                Data curated by PubChem, accessed from: https://pubchem.ncbi.nlm.nih.gov/#tab=compound&query=covid-19%20clinicaltrials (last access date: 30th Apr 2023)""" 
            )         
        ) 
    )
)


# Server---
# Add plotting code within my_widget function within the server function
def server(input, output, session):
    @output
    @render_widget
    def my_widget():
        fig = px.scatter(
            df, x = input.x(), y = input.y(),
            color = "Part_coef_group", 
            hover_name = "Compound name"
        )
        fig.layout.height = 400
        return fig
        
# Combine UI & server into Shiny app
app = App(app_ui, server)


# ***Specify data source***
# --Not the best approach yet but works for now--
# Currently this work-around only suits small to medium-size dataset
# Not ideal for large dataset for sure
# To load file locally use the following code
infile = Path(__file__).parent / "pc_cov_pd.csv"
df = pd.read_csv(infile)

# Then manually paste in csv/txt file below
## file: pc_cov_pd.csv
,Compound name,Molecular weight,Polar surface area,Complexity,Partition coefficients,Heavy atom count,Hydrogen bond donor count,Hydrogen bond acceptor count,Rotatable bond count,Exact mass,Monoisotopic mass,Formal charge,Covalently-bonded unit count,Isotope atom count,Total atom stereocenter count,Defined atom stereocenter count,Undefined atoms stereocenter count,Total bond stereocenter count,Defined bond stereocenter count,Undefined bond stereocenter count,Part_coef_group
0,Calcitriol,416.6,60.7,688.0,5.1,30,3,3,6,416.329,416.329,0,1,0,6,6,0,2,2,0,Larger than 6
1,Ubiquinol,865.4,58.9,1600.0,20.2,63,2,4,31,864.7,864.7,0,1,0,0,0,0,9,9,0,Larger than 6
2,Glutamine,146.14,106.0,146.0,-3.1,10,3,4,4,146.069,146.069,0,1,0,1,1,0,0,0,0,Between -11 and 5
3,Aspirin,180.16,63.6,212.0,1.2,13,1,4,3,180.042,180.042,0,1,0,0,0,0,0,0,0,Between -11 and 5
4,1-Methylnicotinamide,137.16,47.0,136.0,-0.1,10,1,1,1,137.071,137.071,1,1,0,0,0,0,0,0,0,Between -11 and 5
5,Losartan,422.9,92.5,520.0,4.3,30,2,5,8,422.162,422.162,0,1,0,0,0,0,0,0,0,Between -11 and 5
6,Vitamin E,430.7,29.5,503.0,10.7,31,1,2,12,430.381,430.381,0,1,0,3,3,0,0,0,0,Larger than 6
7,Nicotinamide,122.12,56.0,114.0,-0.4,9,1,2,1,122.048,122.048,0,1,0,0,0,0,0,0,0,Between -11 and 5
8,Adenosine,267.24,140.0,335.0,-1.1,19,4,8,2,267.097,267.097,0,1,0,4,4,0,0,0,0,Between -11 and 5
9,Inosine,268.23,129.0,405.0,-1.3,19,4,7,2,268.081,268.081,0,1,0,4,4,0,0,0,0,Between -11 and 5
10,Nicotinamide riboside,255.25,117.0,314.0,-1.8,18,4,5,3,255.098,255.098,1,1,0,4,4,0,0,0,0,Between -11 and 5
11,Dimethyl Fumarate,144.12,52.6,141.0,0.7,10,0,4,4,144.042,144.042,0,1,0,0,0,0,1,1,0,Between -11 and 5
12,Inosinic acid,348.21,176.0,555.0,-3.0,23,5,10,4,348.047,348.047,0,1,0,4,4,0,0,0,0,Between -11 and 5
13,Acetyl-L-carnitine,203.24,66.4,214.0,0.4,14,0,4,5,203.116,203.116,0,1,0,1,1,0,0,0,0,Between -11 and 5
14,Camostat,398.4,137.0,602.0,1.1,29,2,6,9,398.159,398.159,0,1,0,0,0,0,0,0,0,Between -11 and 5
15,Estradiol,272.4,40.5,382.0,4.0,20,2,2,0,272.178,272.178,0,1,0,5,5,0,0,0,0,Between -11 and 5
16,Aspartic Acid,133.1,101.0,133.0,-2.8,9,3,5,3,133.038,133.038,0,1,0,1,1,0,0,0,0,Between -11 and 5
17,Ribavirin,244.2,144.0,304.0,-1.8,17,4,7,3,244.081,244.081,0,1,0,4,4,0,0,0,0,Between -11 and 5
18,Angiotensin (1-7),899.0,380.0,1660.0,-3.0,64,12,14,25,898.466,898.466,0,1,0,8,8,0,0,0,0,Between -11 and 5
19,alpha-Maltose,342.3,190.0,382.0,-4.7,23,8,11,4,342.116,342.116,0,1,0,10,10,0,0,0,0,Between -11 and 5
20,Ambrisentan,378.4,81.5,475.0,3.8,28,1,6,7,378.158,378.158,0,1,0,1,1,0,0,0,0,Between -11 and 5
21,Ciclosporin,1202.6,279.0,2330.0,7.5,85,5,12,15,1201.84,1201.84,0,1,0,12,12,0,1,1,0,Larger than 6
22,Ergocalciferol,396.6,20.2,678.0,7.4,29,1,1,5,396.339,396.339,0,1,0,6,6,0,3,3,0,Larger than 6
23,Docosahexaenoic Acid,328.5,37.3,462.0,6.2,24,1,2,14,328.24,328.24,0,1,0,0,0,0,6,6,0,Larger than 6
24,Tacrolimus,804.0,178.0,1480.0,2.7,57,3,12,7,803.482,803.482,0,1,0,14,14,0,2,2,0,Between -11 and 5
25,Budesonide,430.5,93.1,862.0,2.5,31,2,6,4,430.236,430.236,0,1,0,9,8,1,0,0,0,Between -11 and 5
26,Calcifediol,400.6,40.5,655.0,6.2,29,2,2,6,400.334,400.334,0,1,0,5,5,0,2,2,0,Larger than 6
27,Cepharanthine,606.7,61.9,994.0,6.5,45,0,8,2,606.273,606.273,0,1,0,2,2,0,0,0,0,Larger than 6
28,Cholecalciferol,384.6,20.2,610.0,7.9,28,1,1,6,384.339,384.339,0,1,0,5,5,0,2,2,0,Larger than 6
29,Coenzyme Q10,863.3,52.6,1840.0,19.4,63,0,4,31,862.684,862.684,0,1,0,0,0,0,9,9,0,Larger than 6
30,Tacrolimus monohydrate,822.0,179.0,1480.0,,58,4,13,7,821.493,821.493,0,2,0,14,14,0,2,2,0,Larger than 6
31,Alisporivir,1216.6,279.0,2360.0,7.9,86,5,12,15,1215.86,1215.86,0,1,0,13,13,0,1,1,0,Larger than 6
32,Pacritinib,472.6,68.7,644.0,3.8,35,1,7,4,472.247,472.247,0,1,0,0,0,0,1,1,0,Between -11 and 5
33,Topotecan,421.4,103.0,867.0,0.5,31,2,7,3,421.164,421.164,0,1,0,1,1,0,0,0,0,Between -11 and 5
34,Voclosporin,1214.6,279.0,2380.0,7.9,86,5,12,16,1213.84,1213.84,0,1,0,12,12,0,1,1,0,Larger than 6
35,"(1R,4S,5'S,6R,6'R,8R,10E,12S,13S,14E,16E,20R,21R,24S)-6'-[(2R)-Butan-2-yl]-21,24-dihydroxy-12-[(2R,4S,5S,6S)-5-[(2S,4S,5S,6S)-5-hydroxy-4-methoxy-6-methyloxan-2-yl]oxy-4-methoxy-6-methyloxan-2-yl]oxy-5',11,13,22-tetramethylspiro[3,7,19-trioxatetracyclo[15.6.1.14,8.020,24]pentacosa-10,14,16,22-tetraene-6,2'-oxane]-2-one",875.1,170.0,1680.0,4.1,62,3,14,8,874.508,874.508,0,1,0,20,20,0,3,3,0,Between -11 and 5
36,Oxytocin,1007.2,450.0,1870.0,-2.6,69,12,15,17,1006.44,1006.44,0,1,0,9,9,0,0,0,0,Between -11 and 5
37,"(1R,4S,6R,10E,14E,16E,21R)-6'-butan-2-yl-21,24-dihydroxy-12-[(2R,4S,6S)-5-[(2S,4S,6S)-5-hydroxy-4-methoxy-6-methyloxan-2-yl]oxy-4-methoxy-6-methyloxan-2-yl]oxy-5',11,13,22-tetramethylspiro[3,7,19-trioxatetracyclo[15.6.1.14,8.020,24]pentacosa-10,14,16,22-tetraene-6,2'-oxane]-2-one",875.1,170.0,1680.0,4.1,62,3,14,8,874.508,874.508,0,1,0,20,10,10,3,3,0,Between -11 and 5
38,"Manganese, dichloro((4aS,13aS,17aS,21aS)-1,2,3,4,4a,5,6,12,13,13a,14,15,16,17,17a,18,19,20,21,21a-eicosahydro-7,11-nitrilo-7H-dibenzo(b,H)-5,13,18,21-tetraazacycloheptadecine-kappaN5,kappaN13,kappaN18,kappaN21,kappaN22)-, (pb-7-11-2344'3')-",483.4,61.0,381.0,,29,4,7,0,482.165,482.165,0,4,0,4,4,0,0,0,0,Larger than 6
39,Ivermectin B1a,875.1,170.0,1680.0,4.1,62,3,14,8,874.508,874.508,0,1,0,20,20,0,3,3,0,Between -11 and 5
40,Dipyridamole,504.6,145.0,561.0,0.7,36,4,12,12,504.317,504.317,0,1,0,0,0,0,0,0,0,Between -11 and 5
41,Tetrandrine,622.7,61.9,979.0,6.4,46,0,8,4,622.304,622.304,0,1,0,2,2,0,0,0,0,Larger than 6
42,Sirolimus,914.2,195.0,1760.0,6.0,65,3,13,6,913.555,913.555,0,1,0,15,15,0,4,4,0,Larger than 6
43,Iloprost,360.5,77.8,606.0,2.8,26,3,4,8,360.23,360.23,0,1,0,6,5,1,2,2,0,Between -11 and 5
44,Ramipril,416.5,95.9,619.0,1.4,30,2,6,10,416.231,416.231,0,1,0,5,5,0,0,0,0,Between -11 and 5
45,Prasugrel hydrochloride,409.9,74.8,555.0,,27,1,6,6,409.091,409.091,0,2,0,1,0,1,0,0,0,Larger than 6
46,Uproleselan,1304.5,383.0,1870.0,-2.0,90,9,27,52,1303.72,1303.72,0,1,0,15,15,0,0,0,0,Between -11 and 5
47,MET-enkephalin,573.7,225.0,847.0,-2.1,40,7,9,16,573.226,573.226,0,1,0,3,3,0,0,0,0,Between -11 and 5
48,"(18Z)-1,14-dihydroxy-12-[(E)-1-(4-hydroxy-3-methoxycyclohexyl)prop-1-en-2-yl]-23,25-dimethoxy-13,19,21,27-tetramethyl-17-prop-2-enyl-11,28-dioxa-4-azatricyclo[22.3.1.04,9]octacos-18-ene-2,3,10,16-tetrone",804.0,178.0,1480.0,2.7,57,3,12,7,803.482,803.482,0,1,0,14,0,14,2,2,0,Between -11 and 5
49,"(1R,4S,5'S,6R,6'R,8R,10E,13S,14E,16E,20R,21R,24S)-6'-[(2S)-butan-2-yl]-21,24-dihydroxy-12-[(2R,4S,5S,6S)-5-[(2S,4S,5S,6S)-5-hydroxy-4-methoxy-6-methyloxan-2-yl]oxy-4-methoxy-6-methyloxan-2-yl]oxy-5',11,13,22-tetramethylspiro[3,7,19-trioxatetracyclo[15.6.1.14,8.020,24]pentacosa-10,14,16,22-tetraene-6,2'-oxane]-2-one",875.1,170.0,1680.0,4.1,62,3,14,8,874.508,874.508,0,1,0,20,19,1,3,3,0,Between -11 and 5
50,Plitidepsin,1110.3,285.0,2200.0,5.7,79,4,15,15,1109.63,1109.63,0,1,0,12,12,0,0,0,0,Larger than 6
51,"(10Z,14Z,16Z)-6'-butan-2-yl-21,24-dihydroxy-12-[5-(5-hydroxy-4-methoxy-6-methyloxan-2-yl)oxy-4-methoxy-6-methyloxan-2-yl]oxy-5',11,13,22-tetramethylspiro[3,7,19-trioxatetracyclo[15.6.1.14,8.020,24]pentacosa-10,14,16,22-tetraene-6,2'-oxane]-2-one",875.1,170.0,1680.0,4.1,62,3,14,8,874.508,874.508,0,1,0,20,0,20,3,3,0,Between -11 and 5
52,"(1R,4S,5'S,6R,6'R,8R,12S,13S,20R,21R,24S)-6'-[(2R)-butan-2-yl]-21,24-dihydroxy-12-[(2R,4S,5S,6S)-5-[(2S,4S,5S,6S)-5-hydroxy-4-methoxy-6-methyloxan-2-yl]oxy-4-methoxy-6-methyloxan-2-yl]oxy-5',11,13,22-tetramethylspiro[3,7,19-trioxatetracyclo[15.6.1.14,8.020,24]pentacosa-10,14,16,22-tetraene-6,2'-oxane]-2-one",875.1,170.0,1680.0,4.1,62,3,14,8,874.508,874.508,0,1,0,20,20,0,3,0,3,Between -11 and 5
53,Rubramin,1355.4,476.0,3220.0,,93,9,21,16,1354.57,1354.57,-3,3,0,14,14,0,3,3,0,Larger than 6
54,"(1R,4S,5'S,6R,6'S,8R,10E,12S,13S,14E,16E,20R,21R,24S)-6'-[(2S)-butan-2-yl]-21,24-dihydroxy-12-[(2R,4S,5S,6S)-5-[(2S,4S,5S,6S)-5-hydroxy-4-methoxy-6-methyloxan-2-yl]oxy-4-methoxy-6-methyloxan-2-yl]oxy-5',11,13,22-tetramethylspiro[3,7,19-trioxatetracyclo[15.6.1.14,8.020,24]pentacosa-10,14,16,22-tetraene-6,2'-oxane]-2-one",875.1,170.0,1680.0,4.1,62,3,14,8,874.508,874.508,0,1,0,20,20,0,3,3,0,Between -11 and 5
55,Zidovudine,267.24,93.2,484.0,0.0,19,2,6,3,267.097,267.097,0,1,0,3,3,0,0,0,0,Between -11 and 5
56,Resveratrol,228.24,60.7,246.0,3.1,17,3,3,2,228.079,228.079,0,1,0,0,0,0,1,1,0,Between -11 and 5
57,Curcumin,368.4,93.1,507.0,3.2,27,2,6,8,368.126,368.126,0,1,0,0,0,0,2,2,0,Between -11 and 5
58,"6'-Butan-2-yl-21,24-dihydroxy-12-[5-(5-hydroxy-4-methoxy-6-methyloxan-2-yl)oxy-4-methoxy-6-methyloxan-2-yl]oxy-5',11,13,22-tetramethylspiro[3,7,19-trioxatetracyclo[15.6.1.14,8.020,24]pentacosa-10,14,16,22-tetraene-6,2'-oxane]-2-one",875.1,170.0,1680.0,4.1,62,3,14,8,874.508,874.508,0,1,0,20,0,20,3,0,3,Between -11 and 5
59,Regorafenib,482.8,92.4,686.0,4.2,33,3,8,5,482.077,482.077,0,1,0,0,0,0,0,0,0,Between -11 and 5
60,Pharmakon1600-01300027,875.1,170.0,1680.0,4.1,62,3,14,8,874.508,874.508,0,1,0,20,17,3,3,3,0,Between -11 and 5
61,"(1R,4S,5'S,6R,6'R,8R,10Z,12S,13S,14Z,20R,21R,24S)-6'-[(2S)-butan-2-yl]-21,24-dihydroxy-12-[(2R,4S,5S,6S)-5-[(2S,4S,5S,6S)-5-hydroxy-4-methoxy-6-methyloxan-2-yl]oxy-4-methoxy-6-methyloxan-2-yl]oxy-5',11,13,22-tetramethylspiro[3,7,19-trioxatetracyclo[15.6.1.14,8.020,24]pentacosa-10,14,16,22-tetraene-6,2'-oxane]-2-one",875.1,170.0,1680.0,4.1,62,3,14,8,874.508,874.508,0,1,0,20,20,0,3,2,1,Between -11 and 5
62,"(1R,4S,5'S,6R,6'R,8R,10E,12S,13S,14E,16E,20R,21R,24S)-6'-[(2S)-butan-2-yl]-21,24-dihydroxy-12-[(2R,4R,5S,6S)-5-[(2S,4R,5S,6S)-5-hydroxy-4-methoxy-6-methyloxan-2-yl]oxy-4-methoxy-6-methyloxan-2-yl]oxy-5',11,13,22-tetramethylspiro[3,7,19-trioxatetracyclo[15.6.1.14,8.020,24]pentacosa-10,14,16,22-tetraene-6,2'-oxane]-2-one",875.1,170.0,1680.0,4.1,62,3,14,8,874.508,874.508,0,1,0,20,20,0,3,3,0,Between -11 and 5
63,Methylcobalamin,1345.4,449.0,3160.0,,92,10,20,26,1344.6,1344.6,0,3,0,14,13,1,3,3,0,Larger than 6
64,Spironolactone,416.6,85.7,818.0,2.9,29,0,5,2,416.202,416.202,0,1,0,7,7,0,0,0,0,Between -11 and 5
65,Prednisone,358.4,91.7,764.0,1.5,26,2,5,2,358.178,358.178,0,1,0,6,6,0,0,0,0,Between -11 and 5
66,Quercetin,302.23,127.0,488.0,1.5,22,5,7,1,302.043,302.043,0,1,0,0,0,0,0,0,0,Between -11 and 5
67,Calderol,418.7,41.5,655.0,,30,3,3,6,418.345,418.345,0,2,0,5,5,0,2,2,0,Larger than 6
68,Dactolisib,469.5,73.1,872.0,5.2,36,0,4,3,469.19,469.19,0,1,0,0,0,0,0,0,0,Larger than 6
69,Enzalutamide,464.4,109.0,839.0,3.6,32,1,8,3,464.093,464.093,0,1,0,0,0,0,0,0,0,Between -11 and 5
70,Selinexor,443.3,97.6,621.0,3.0,31,2,12,5,443.093,443.093,0,1,0,0,0,0,1,1,0,Between -11 and 5
71,Chlorpromazine,318.9,31.8,339.0,5.2,21,0,3,4,318.096,318.096,0,1,0,0,0,0,0,0,0,Larger than 6
72,Itraconazolum [Latin],705.6,101.0,1120.0,5.7,49,0,9,11,704.239,704.239,0,1,0,3,0,3,0,0,0,Larger than 6
73,Loratadine,382.9,42.4,569.0,5.2,27,0,3,2,382.145,382.145,0,1,0,0,0,0,0,0,0,Larger than 6
74,Nifedipine,346.3,110.0,608.0,2.2,25,1,7,5,346.116,346.116,0,1,0,0,0,0,0,0,0,Between -11 and 5
75,Prazosin,383.4,107.0,544.0,2.0,28,1,8,4,383.159,383.159,0,1,0,0,0,0,0,0,0,Between -11 and 5
76,Alizarin,240.21,74.6,378.0,3.2,18,2,4,0,240.042,240.042,0,1,0,0,0,0,0,0,0,Between -11 and 5
77,Methylprednisolone,374.5,94.8,754.0,1.9,27,3,5,2,374.209,374.209,0,1,0,8,8,0,0,0,0,Between -11 and 5
78,Chloroquine sulfate,418.0,111.0,390.0,,27,3,7,8,417.149,417.149,0,2,0,1,0,1,0,0,0,Larger than 6
79,Bardoxolone methyl,505.7,84.2,1210.0,6.7,37,0,5,2,505.319,505.319,0,1,0,7,7,0,0,0,0,Larger than 6
80,"(1R,4S,5'S,6R,6'R,8R,12S,13S,20R,21R,24S)-6'-[(2S)-butan-2-yl]-21,24-dihydroxy-12-[(2R,4S,5S,6S)-5-[(2S,4S,5S,6S)-5-hydroxy-4-methoxy-6-methyloxan-2-yl]oxy-4-methoxy-6-methyloxan-2-yl]oxy-5',11,13,22-tetramethylspiro[3,7,19-trioxatetracyclo[15.6.1.14,8.020,24]pentacosa-10,14,16,22-tetraene-6,2'-oxane]-2-one",875.1,170.0,1680.0,4.1,62,3,14,8,874.508,874.508,0,1,0,20,20,0,3,0,3,Between -11 and 5
81,Heartgard-30,1736.2,340.0,3340.0,,123,6,28,15,1736.0,1735.0,0,2,0,39,34,5,6,6,0,Larger than 6
82,Ivermectine 100 microg/mL in Acetonitrile,1736.2,340.0,3340.0,,123,6,28,15,1736.0,1735.0,0,2,0,39,39,0,6,6,0,Larger than 6
83,Duvelisib,416.9,86.8,668.0,4.1,30,2,5,4,416.115,416.115,0,1,0,1,1,0,0,0,0,Between -11 and 5
84,"(4S,5'S,6R,6'R,8R,10E,12S,13S,14E,16E,20R,21R,24S)-6'-[(2S)-butan-2-yl]-21,24-dihydroxy-12-[(2R,4S,5S,6S)-5-[(2S,4S,5S,6S)-5-hydroxy-4-methoxy-6-methyloxan-2-yl]oxy-4-methoxy-6-methyloxan-2-yl]oxy-5',11,13,22-tetramethylspiro[3,7,19-trioxatetracyclo[15.6.1.14,8.020,24]pentacosa-10,14,16,22-tetraene-6,2'-oxane]-2-one",875.1,170.0,1680.0,4.1,62,3,14,8,874.508,874.508,0,1,0,20,19,1,3,3,0,Between -11 and 5
85,Zotatifin,487.5,108.0,819.0,2.4,36,2,8,6,487.211,487.211,0,1,0,5,5,0,0,0,0,Between -11 and 5
86,MaxEPA,645.0,74.6,874.0,,47,2,4,28,644.48,644.48,0,2,0,0,0,0,11,11,0,Larger than 6
87,"(1R,4S,5'S,6R,6'R,8R,12S,13S,14E,16E,20R,21R,24S)-6'-[(2S)-butan-2-yl]-21,24-dihydroxy-12-[(2R,4S,5S,6S)-5-[(2S,4S,5S,6S)-5-hydroxy-4-methoxy-6-methyloxan-2-yl]oxy-4-methoxy-6-methyloxan-2-yl]oxy-5',11,13,22-tetramethylspiro[3,7,19-trioxatetracyclo[15.6.1.14,8.020,24]pentacosa-10,14,16,22-tetraene-6,2'-oxane]-2-one",875.1,170.0,1680.0,4.1,62,3,14,8,874.508,874.508,0,1,0,20,20,0,3,2,1,Between -11 and 5
88,"(1S,4S,5'S,6R,6'R,8R,10E,12S,13S,14E,20R,21R,24S)-6'-[(2S)-butan-2-yl]-21,24-dihydroxy-12-[(2R,4S,5S,6S)-5-[(2S,4S,5S,6S)-5-hydroxy-4-methoxy-6-methyloxan-2-yl]oxy-4-methoxy-6-methyloxan-2-yl]oxy-5',11,13,22-tetramethylspiro[3,7,19-trioxatetracyclo[15.6.1.14,8.020,24]pentacosa-10,14,16,22-tetraene-6,2'-oxane]-2-one",875.1,170.0,1680.0,4.1,62,3,14,8,874.508,874.508,0,1,0,20,20,0,3,2,1,Between -11 and 5
89,Indomethacin,357.8,68.5,506.0,4.3,25,1,4,4,357.077,357.077,0,1,0,0,0,0,0,0,0,Between -11 and 5
90,Methylene Blue,319.9,43.9,483.0,,21,0,4,1,319.091,319.091,0,2,0,0,0,0,0,0,0,Larger than 6
91,Bromocresol green,698.0,92.2,742.0,6.5,31,2,5,2,697.725,693.73,0,1,0,0,0,0,0,0,0,Larger than 6
92,Hydroxychloroquine sulfate,434.0,131.0,413.0,,28,4,8,9,433.144,433.144,0,2,0,1,0,1,0,0,0,Larger than 6
93,Ritonavir,720.9,202.0,1040.0,6.0,50,4,9,18,720.313,720.313,0,1,0,4,4,0,0,0,0,Larger than 6
94,Fisetin,286.24,107.0,459.0,2.0,21,4,6,1,286.048,286.048,0,1,0,0,0,0,0,0,0,Between -11 and 5
95,Palbociclib,447.5,103.0,775.0,1.8,33,2,8,5,447.238,447.238,0,1,0,0,0,0,0,0,0,Between -11 and 5
96,Ruxolitinib,306.4,83.2,453.0,2.1,23,1,4,4,306.159,306.159,0,1,0,1,1,0,0,0,0,Between -11 and 5
97,Sildenafil Citrate,666.7,250.0,1070.0,,46,5,15,12,666.232,666.232,0,2,0,0,0,0,0,0,0,Larger than 6
98,Nintedanib,539.6,102.0,892.0,4.3,40,2,7,8,539.253,539.253,0,1,0,0,0,0,0,0,0,Between -11 and 5
99,Melatonin,232.28,54.1,270.0,0.8,17,2,2,4,232.121,232.121,0,1,0,0,0,0,0,0,0,Between -11 and 5
100,Carbamazepine,236.27,46.3,326.0,2.5,18,1,1,0,236.095,236.095,0,1,0,0,0,0,0,0,0,Between -11 and 5
101,Clofazimine,473.4,40.0,829.0,7.1,33,1,4,4,472.122,472.122,0,1,0,0,0,0,0,0,0,Larger than 6
102,Diphenhydramine,255.35,12.5,211.0,3.3,19,0,2,6,255.162,255.162,0,1,0,0,0,0,0,0,0,Between -11 and 5
103,Thalidomide,258.23,83.6,449.0,0.3,19,1,4,1,258.064,258.064,0,1,0,1,0,1,0,0,0,Between -11 and 5
104,Hydrocortisone,362.5,94.8,684.0,1.6,26,3,5,2,362.209,362.209,0,1,0,7,7,0,0,0,0,Between -11 and 5
105,Progesterone,314.5,34.1,589.0,3.9,23,0,2,1,314.225,314.225,0,1,0,6,6,0,0,0,0,Between -11 and 5
106,o-Aminoazotoluene,225.29,50.7,264.0,3.7,17,1,3,2,225.127,225.127,0,1,0,0,0,0,0,0,0,Between -11 and 5
107,Clonitralid,388.2,141.0,414.0,,25,4,6,3,387.039,387.039,0,2,0,0,0,0,0,0,0,Larger than 6
108,Pimozide,461.5,35.6,632.0,6.3,34,1,4,7,461.228,461.228,0,1,0,0,0,0,0,0,0,Larger than 6
109,Nitazoxanide,307.28,142.0,428.0,2.0,21,1,7,4,307.026,307.026,0,1,0,0,0,0,0,0,0,Between -11 and 5
110,Deferoxamine hydrochloride,597.1,206.0,739.0,,40,7,9,23,596.33,596.33,0,2,0,0,0,0,0,0,0,Larger than 6
111,Chloroquine monophosphate,417.9,106.0,359.0,,27,4,7,8,417.158,417.158,0,2,0,1,0,1,0,0,0,Larger than 6
112,Retinol,286.5,20.2,496.0,5.7,21,1,1,5,286.23,286.23,0,1,0,0,0,0,4,4,0,Larger than 6
113,Melphalan,305.2,66.6,265.0,-0.5,19,2,4,8,304.075,304.075,0,1,0,1,1,0,0,0,0,Between -11 and 5
114,Dasatinib,488.0,135.0,642.0,3.6,33,3,9,7,487.156,487.156,0,1,0,0,0,0,0,0,0,Between -11 and 5
115,Masitinib,498.6,102.0,696.0,4.3,36,2,7,7,498.22,498.22,0,1,0,0,0,0,0,0,0,Between -11 and 5
116,Acrivastine and pseudoephedrine hydrochloride,550.1,85.7,635.0,,39,4,6,9,549.276,549.276,0,3,0,2,2,0,2,2,0,Larger than 6
117,"4-(carboxymethyl)-2-((R)-1-(2-(2,5-dichlorobenzamido)acetamido)-3-methylbutyl)-6-oxo-1,3,2-dioxaborinane-4-carboxylic acid",517.1,168.0,815.0,,34,4,9,10,516.087,516.087,0,1,0,2,1,1,0,0,0,Larger than 6
118,Remdesivir,602.6,204.0,1010.0,1.9,42,4,13,14,602.225,602.225,0,1,0,6,6,0,0,0,0,Between -11 and 5
119,"Ivermectin B1a, epi-",875.1,170.0,1680.0,4.1,62,3,14,8,874.508,874.508,0,1,0,20,20,0,3,3,0,Between -11 and 5
120,Hydroxocobalamin,1270.4,452.0,3140.0,-3.4,90,9,19,26,1269.63,1269.63,-2,1,0,14,0,14,3,3,0,Between -11 and 5
121,Zanubrutinib,471.5,103.0,756.0,3.5,35,2,5,6,471.227,471.227,0,1,0,1,1,0,0,0,0,Between -11 and 5
122,Gamma-Aminobutyric Acid,103.12,63.3,62.7,-3.2,7,2,3,3,103.063,103.063,0,1,0,0,0,0,0,0,0,Between -11 and 5
123,Caffeine,194.19,58.4,293.0,-0.1,14,0,3,0,194.08,194.08,0,1,0,0,0,0,0,0,0,Between -11 and 5
124,Dapsone,248.3,94.6,306.0,1.0,17,2,4,2,248.062,248.062,0,1,0,0,0,0,0,0,0,Between -11 and 5
125,Leflunomide,270.21,55.1,327.0,2.5,19,1,6,2,270.062,270.062,0,1,0,0,0,0,0,0,0,Between -11 and 5
126,Raloxifene,473.6,98.2,655.0,6.1,34,2,6,7,473.166,473.166,0,1,0,0,0,0,0,0,0,Larger than 6
127,Imatinib,493.6,86.3,706.0,3.5,37,2,7,7,493.259,493.259,0,1,0,0,0,0,0,0,0,Between -11 and 5
128,Dexamethasone,392.5,94.8,805.0,1.9,28,3,6,2,392.2,392.2,0,1,0,8,8,0,0,0,0,Between -11 and 5
129,Colchicine,399.4,83.1,740.0,1.0,29,1,6,5,399.168,399.168,0,1,0,1,1,0,0,0,0,Between -11 and 5
130,Estradiol cypionate,396.6,46.5,597.0,7.1,29,1,3,5,396.266,396.266,0,1,0,5,5,0,0,0,0,Larger than 6
131,Cetylpyridinium Chloride,340.0,3.9,208.0,,23,0,1,15,339.269,339.269,0,2,0,0,0,0,0,0,0,Larger than 6
132,Plerixafor,502.8,78.7,456.0,0.0,36,6,8,4,502.447,502.447,0,1,0,0,0,0,0,0,0,Between -11 and 5
133,Telmisartan,514.6,72.9,831.0,6.9,39,1,4,7,514.237,514.237,0,1,0,0,0,0,0,0,0,Larger than 6
134,Sorafenib,464.8,92.4,646.0,4.1,32,3,7,5,464.086,464.086,0,1,0,0,0,0,0,0,0,Between -11 and 5
135,Bacitracin zinc,1488.1,552.0,2950.0,,101,16,21,31,1485.68,1485.68,0,2,0,16,0,16,0,0,0,Larger than 6
136,Hymecromone,176.17,46.5,257.0,1.9,13,1,3,0,176.047,176.047,0,1,0,0,0,0,0,0,0,Between -11 and 5
137,Cob(II)alamin,1329.3,452.0,3150.0,,91,9,19,16,1328.56,1328.56,-2,2,0,14,14,0,3,3,0,Larger than 6
138,Tofacitinib,312.37,88.9,488.0,1.5,23,1,5,3,312.17,312.17,0,1,0,2,2,0,0,0,0,Between -11 and 5
139,Elsulfavirine,629.3,134.0,977.0,5.2,37,2,7,8,626.943,626.943,0,1,0,0,0,0,0,0,0,Larger than 6
140,Fostamatinib,580.5,187.0,904.0,1.6,40,4,15,10,580.148,580.148,0,1,0,0,0,0,0,0,0,Between -11 and 5
141,3-(4-chlorophenyl)-N-(pyridin-4-ylmethyl)adamantane-1-carboxamide,380.9,42.0,551.0,4.7,27,1,2,4,380.166,380.166,0,1,0,2,0,2,0,0,0,Between -11 and 5
142,Adenosylcobalamin,1579.6,571.0,3730.0,,109,12,27,17,1578.66,1578.66,-3,3,0,18,18,0,3,3,0,Larger than 6
143,Solnatide,1923.1,861.0,4170.0,-10.7,134,27,32,27,1921.81,1921.81,0,1,0,16,16,0,0,0,0,Smaller than -10
144,Subasumstat,578.1,193.0,942.0,3.5,38,4,11,8,577.122,577.122,0,1,0,4,4,0,0,0,0,Between -11 and 5
145,Lactoferrin,3125.8,1360.0,7330.0,6.8,219,51,65,108,3124.68,3123.68,0,1,0,26,26,0,0,0,0,Larger than 6
146,Compstatin 40,1789.1,692.0,3770.0,-2.1,126,21,23,30,1787.84,1787.84,0,1,0,15,15,0,0,0,0,Between -11 and 5
147,Celecoxib,381.4,86.4,577.0,3.4,26,1,7,3,381.076,381.076,0,1,0,0,0,0,0,0,0,Between -11 and 5
148,Eucalyptol,154.25,9.2,164.0,2.5,11,0,1,0,154.136,154.136,0,1,0,0,0,0,0,0,0,Between -11 and 5
149,Disulfiram,296.5,121.0,201.0,3.9,16,0,4,7,296.051,296.051,0,1,0,0,0,0,0,0,0,Between -11 and 5
150,Ibuprofen,206.28,37.3,203.0,3.5,15,1,2,4,206.131,206.131,0,1,0,1,0,1,0,0,0,Between -11 and 5
151,Niclosamide,327.12,95.2,404.0,4.0,21,2,4,2,325.986,325.986,0,1,0,0,0,0,0,0,0,Between -11 and 5
152,Simvastatin,418.6,72.8,706.0,4.7,30,1,5,7,418.272,418.272,0,1,0,7,7,0,0,0,0,Between -11 and 5
153,Methotrexate,454.4,211.0,704.0,-1.8,33,5,12,9,454.171,454.171,0,1,0,1,1,0,0,0,0,Between -11 and 5
154,Angiotensin II,1046.2,409.0,1980.0,-1.7,75,13,15,29,1045.53,1045.53,0,1,0,9,9,0,0,0,0,Between -11 and 5
155,Lenalidomide,259.26,92.5,437.0,-0.5,19,2,4,1,259.096,259.096,0,1,0,1,0,1,0,0,0,Between -11 and 5
156,Aspartyl-alanyl-diketopiperazine,186.17,95.5,263.0,-1.4,13,3,4,2,186.064,186.064,0,1,0,2,2,0,0,0,0,Between -11 and 5
157,Silmitasertib,349.8,75.1,491.0,4.4,25,2,5,3,349.062,349.062,0,1,0,0,0,0,0,0,0,Between -11 and 5
158,Ibrutinib,440.5,99.2,678.0,3.6,33,1,6,5,440.196,440.196,0,1,0,1,1,0,0,0,0,Between -11 and 5
159,Bemcentinib,506.6,97.8,775.0,5.5,38,2,7,4,506.291,506.291,0,1,0,1,1,0,0,0,0,Larger than 6
160,Teriflunomide,270.21,73.1,426.0,3.3,19,2,6,2,270.062,270.062,0,1,0,0,0,0,1,1,0,Between -11 and 5
161,Tazemetostat,572.7,83.1,992.0,4.2,42,2,6,9,572.336,572.336,0,1,0,0,0,0,0,0,0,Between -11 and 5
162,Acalabrutinib,465.5,119.0,845.0,3.0,35,2,6,4,465.191,465.191,0,1,0,1,1,0,0,0,0,Between -11 and 5
163,Ifenprodil,325.4,43.7,353.0,3.9,24,2,3,5,325.204,325.204,0,1,0,2,0,2,0,0,0,Between -11 and 5
164,Ursodeoxycholic acid,392.6,77.8,605.0,4.9,28,3,4,4,392.293,392.293,0,1,0,10,10,0,0,0,0,Between -11 and 5
165,Metoprolol succinate,652.8,176.0,308.0,,46,6,12,21,652.393,652.393,0,3,0,2,0,2,0,0,0,Larger than 6
166,Merimepodib,452.5,124.0,652.0,2.1,33,3,7,8,452.17,452.17,0,1,0,1,1,0,0,0,0,Between -11 and 5
167,Fludrocortisone acetate,422.5,101.0,838.0,1.7,30,2,7,4,422.21,422.21,0,1,0,7,7,0,0,0,0,Between -11 and 5
168,Triazavirin,228.19,141.0,435.0,0.4,15,1,6,1,228.007,228.007,0,1,0,0,0,0,0,0,0,Between -11 and 5
169,Ceftriaxone sodium,598.6,297.0,1120.0,,38,2,14,7,598.01,598.01,0,3,0,2,2,0,1,1,0,Larger than 6
170,Rivaroxaban,435.9,116.0,645.0,2.5,29,1,6,5,435.066,435.066,0,1,0,1,1,0,0,0,0,Between -11 and 5
171,Ventolin,337.39,156.0,309.0,,22,6,8,5,337.12,337.12,0,2,0,1,0,1,0,0,0,Larger than 6
172,Apixaban,459.5,111.0,777.0,2.2,34,1,5,5,459.191,459.191,0,1,0,0,0,0,0,0,0,Between -11 and 5
173,Ceftriaxone disodium salt hemiheptahydrate,1323.2,601.0,1120.0,,83,11,35,14,1322.09,1322.09,0,13,0,4,4,0,2,2,0,Larger than 6
174,Vadadustat,306.7,99.5,393.0,2.5,21,3,5,4,306.041,306.041,0,1,0,0,0,0,0,0,0,Between -11 and 5
175,Carbohydrate moiety of bromelain,1026.9,483.0,1680.0,-11.6,70,18,29,16,1026.38,1026.38,0,1,0,29,28,1,0,0,0,Smaller than -10
176,Siponimod fumarate,1149.3,199.0,896.0,,82,4,20,20,1148.53,1148.53,0,3,0,0,0,0,3,3,0,Larger than 6
177,(s)-1-(3-Chloro-4-fluorophenyl)ethanamine hydrochloride,210.07,26.0,131.0,,12,2,2,1,209.017,209.017,0,2,0,1,1,0,0,0,0,Larger than 6
178,Prezcobix,1323.7,343.0,1980.0,,92,6,19,32,1322.59,1322.59,0,2,0,8,8,0,0,0,0,Larger than 6
179,"[(2R,3R,4R,5R)-5-(4-amino-5-deuteriopyrrolo[2,1-f][1,2,4]triazin-7-yl)-5-cyano-3,4-bis(2-methylpropanoyloxy)oxolan-2-yl]methyl 2-methylpropanoate;hydrobromide",583.4,168.0,887.0,,37,2,11,11,582.155,582.155,0,2,1,4,4,0,0,0,0,Larger than 6
180,Berberine,336.4,40.8,488.0,3.6,25,0,4,2,336.124,336.124,1,1,0,0,0,0,0,0,0,Between -11 and 5
181,Cyproheptadine,287.4,3.2,423.0,4.7,22,0,1,0,287.167,287.167,0,1,0,0,0,0,0,0,0,Between -11 and 5
182,Doxazosin,451.5,112.0,678.0,2.5,33,1,9,4,451.186,451.186,0,1,0,1,0,1,0,0,0,Between -11 and 5
183,Fluconazole,306.27,81.6,358.0,0.4,22,1,7,5,306.104,306.104,0,1,0,0,0,0,0,0,0,Between -11 and 5
184,Irbesartan,428.5,87.1,682.0,4.1,32,1,5,7,428.232,428.232,0,1,0,0,0,0,0,0,0,Between -11 and 5
185,Sulfamethoxazole,253.28,107.0,346.0,0.9,17,2,6,3,253.052,253.052,0,1,0,0,0,0,0,0,0,Between -11 and 5
186,D-Glucose,180.16,110.0,151.0,-2.6,12,5,6,1,180.063,180.063,0,1,0,5,4,1,0,0,0,Between -11 and 5
187,N-Vinyl-2-pyrrolidone,111.14,20.3,120.0,0.4,8,0,1,1,111.068,111.068,0,1,0,0,0,0,0,0,0,Between -11 and 5
188,Trimetazidine,266.34,43.0,259.0,1.0,19,1,5,5,266.163,266.163,0,1,0,0,0,0,0,0,0,Between -11 and 5
189,Estetrol,304.4,80.9,441.0,1.5,22,4,4,0,304.167,304.167,0,1,0,7,7,0,0,0,0,Between -11 and 5
190,Deferoxamine mesylate,656.8,269.0,832.0,,44,7,12,23,656.341,656.341,0,2,0,0,0,0,0,0,0,Larger than 6
191,Tramadol Hydrochloride,299.83,32.7,282.0,,20,2,3,4,299.165,299.165,0,2,0,2,2,0,0,0,0,Larger than 6
192,Nebivolol,405.4,71.0,483.0,3.0,29,3,7,6,405.175,405.175,0,1,0,4,0,4,0,0,0,Between -11 and 5
193,Argatroban monohydrate,526.7,190.0,887.0,,36,6,9,9,526.257,526.257,0,2,0,4,3,1,0,0,0,Larger than 6
194,Sulfamethoxazole and trimethoprim,543.6,212.0,653.0,,38,4,13,8,543.19,543.19,0,2,0,0,0,0,0,0,0,Larger than 6
195,Zinc Gluconate,455.7,283.0,165.0,,27,10,14,8,454.03,454.03,0,3,0,8,8,0,0,0,0,Larger than 6
196,"Zinc;(2R,3S,4R,5R)-2,3,4,5,6-pentahydroxyhexanoate",455.7,283.0,385.0,,27,10,14,8,454.03,454.03,0,3,0,8,8,0,0,0,0,Larger than 6
197,(+)-Mefloquine,378.31,45.2,483.0,3.6,26,2,9,2,378.117,378.117,0,1,0,2,2,0,0,0,0,Between -11 and 5
198,"(2S,5R,6R)-6-[[(2S)-2-[(4-ethyl-2,3-dioxo-piperazine-1-carbonyl)amino]-2-phenyl-propanoyl]amino]-3,3-dimethyl-7-oxo-4-thia-1-azabicyclo[3.2.0]heptane-2-carboxylic acid; (2S,3S,5R)-3-methyl-4,4,7-trioxo-3-(triazol-1-ylmethyl)-4$l^{6}-thia-1-azabicyclo[3.2.0]heptane-2-carboxylic acid",831.9,313.0,1600.0,,57,4,15,9,831.232,831.232,0,2,0,7,7,0,0,0,0,Larger than 6
199,Favipiravir,157.1,84.6,282.0,-0.6,11,2,4,1,157.029,157.029,0,1,0,0,0,0,0,0,0,Between -11 and 5
200,Maraviroc,513.7,63.0,751.0,5.1,37,1,6,8,513.328,513.328,0,1,0,3,3,0,0,0,0,Larger than 6
201,Toremifene,406.0,12.5,483.0,7.2,29,0,2,9,405.186,405.186,0,1,0,0,0,0,1,1,0,Larger than 6
202,"(3-methyl-2,4-dioxo-3,4-dihydroquinazolin-1(2H)-yl)acetic acid",234.21,77.9,369.0,0.4,17,1,4,2,234.064,234.064,0,1,0,0,0,0,0,0,0,Between -11 and 5
203,Enalapril maleate,492.5,171.0,638.0,,35,4,10,12,492.211,492.211,0,2,0,3,3,0,1,1,0,Larger than 6
204,Naltrexone hydrochloride,377.9,70.0,621.0,,26,3,5,2,377.139,377.139,0,2,0,4,4,0,0,0,0,Larger than 6
205,Chlorhexidine Gluconate,897.8,455.0,819.0,,60,18,16,23,896.32,896.32,0,3,0,8,8,0,2,2,0,Larger than 6
206,Piperacillin/tazobactam,817.9,313.0,1550.0,,56,4,15,9,817.216,817.216,0,2,0,7,7,0,0,0,0,Larger than 6
207,3-[5-(azetidine-1-carbonyl)pyrazin-2-yl]oxy-5-[(2S)-1-methoxypropan-2-yl]oxy-N-(5-methylpyrazin-2-yl)benzamide,478.5,129.0,710.0,1.3,35,1,9,9,478.196,478.196,0,1,0,1,1,0,0,0,0,Between -11 and 5
208,Degarelix,1632.3,513.0,3390.0,3.5,117,17,18,41,1630.75,1630.75,0,1,0,11,11,0,0,0,0,Between -11 and 5
209,(3S)-3-amino-6-(diaminomethylideneazaniumyl)hex-1-en-2-olate,172.23,115.0,174.0,-0.8,12,4,2,5,172.132,172.132,0,1,0,1,1,0,0,0,0,Between -11 and 5
210,Sivelestat sodium,528.5,154.0,738.0,,35,6,12,9,528.139,528.139,0,6,0,0,0,0,0,0,0,Larger than 6
211,CID 23679441,576.6,291.0,1120.0,,37,3,13,8,576.028,576.028,0,2,0,2,2,0,1,1,0,Larger than 6
212,Masitinib mesylate,594.8,164.0,788.0,,41,3,10,7,594.208,594.208,0,2,0,0,0,0,0,0,0,Larger than 6
213,"[(1S,4R,6S,7Z,18R)-4-(cyclopropylsulfonylcarbamoyl)-14-[(2-methylpropan-2-yl)oxycarbonylamino]-2,15-dioxo-3,16-diazatricyclo[14.3.0.04,6]nonadec-7-en-18-yl] 4-fluoro-1,3-dihydroisoindole-2-carboxylate",731.8,189.0,1530.0,3.3,51,3,10,8,731.3,731.3,0,1,0,5,4,1,1,1,0,Between -11 and 5
214,Natrii chloridi solutio composita,309.11,63.4,24.8,,13,4,8,0,307.852,307.852,0,10,0,0,0,0,0,0,0,Larger than 6
215,CID 87060529,577.6,288.0,1110.0,,37,4,13,8,577.036,577.036,0,2,0,2,2,0,1,1,0,Larger than 6
216,"(2S,3R)-3-[(2-aminopyridin-4-yl)methyl]-1-[[(1R)-1-cyclooctylethyl]carbamoyl]-4-oxoazetidine-2-carboxylic acid",402.5,126.0,606.0,3.7,29,3,6,5,402.227,402.227,0,1,0,3,3,0,0,0,0,Between -11 and 5
217,Sinapultide acetate,2529.5,813.0,4910.0,,178,28,30,93,2528.85,2527.85,0,2,0,21,21,0,0,0,0,Larger than 6
218,Bemnifosbuvir hemisulfate,1261.1,453.0,1000.0,,85,10,32,24,1260.4,1260.4,0,3,0,12,12,0,0,0,0,Larger than 6
219,Reamberin,357.27,193.0,216.0,,23,6,10,7,357.101,357.101,0,4,0,4,4,0,0,0,0,Larger than 6
220,Pomotrelvir,455.9,127.0,779.0,3.1,32,4,4,8,455.172,455.172,0,1,0,3,3,0,0,0,0,Between -11 and 5
221,Selenious acid,128.99,57.5,26.3,,4,2,3,0,129.917,129.917,0,1,0,0,0,0,0,0,0,Larger than 6
222,Spermidine,145.25,64.099,56.8,-1.0,10,3,3,7,145.158,145.158,0,1,0,0,0,0,0,0,0,Between -11 and 5
223,Salbutamol,239.31,72.7,227.0,0.3,17,4,4,5,239.152,239.152,0,1,0,1,0,1,0,0,0,Between -11 and 5
224,Bromhexine,376.13,29.3,256.0,4.3,18,1,2,3,375.997,373.999,0,1,0,0,0,0,0,0,0,Between -11 and 5
225,Ebselen,274.19,20.3,275.0,,16,0,1,1,274.985,274.985,0,1,0,0,0,0,0,0,0,Larger than 6
226,Fluoxetine,309.33,21.3,308.0,4.0,22,1,5,6,309.134,309.134,0,1,0,1,0,1,0,0,0,Between -11 and 5
227,Ketotifen,309.4,48.6,476.0,3.2,22,0,3,0,309.119,309.119,0,1,0,0,0,0,0,0,0,Between -11 and 5
228,Mefloquine,378.31,45.2,483.0,3.6,26,2,9,2,378.117,378.117,0,1,0,2,0,2,0,0,0,Between -11 and 5
229,Midazolam,325.8,30.2,471.0,2.5,23,0,3,1,325.078,325.078,0,1,0,0,0,0,0,0,0,Between -11 and 5
230,Nitroglycerin,227.09,165.0,219.0,1.6,15,0,9,5,227.003,227.003,0,1,0,0,0,0,0,0,0,Between -11 and 5
231,Quetiapine,383.5,73.6,496.0,2.1,27,1,5,6,383.167,383.167,0,1,0,0,0,0,0,0,0,Between -11 and 5
232,Liothyronine,650.97,92.8,402.0,1.7,23,3,5,5,650.79,650.79,0,1,0,1,1,0,0,0,0,Between -11 and 5
233,Pyridostigmine bromide,261.12,33.4,183.0,,14,0,3,2,260.016,260.016,0,2,0,0,0,0,0,0,0,Larger than 6
234,Psilocybine,284.25,85.8,347.0,-1.6,19,3,5,5,284.093,284.093,0,1,0,0,0,0,0,0,0,Between -11 and 5
235,Luminol,177.16,84.2,254.0,0.3,13,3,3,0,177.054,177.054,0,1,0,0,0,0,0,0,0,Between -11 and 5
236,Acetylcysteine,163.2,67.4,148.0,0.4,10,3,4,3,163.03,163.03,0,1,0,1,1,0,0,0,0,Between -11 and 5
237,Cyproheptadine hydrochloride,323.9,3.2,423.0,,23,1,1,0,323.144,323.144,0,2,0,0,0,0,0,0,0,Larger than 6
238,Carmoisine,502.4,176.0,854.0,,33,1,9,2,501.988,501.988,0,3,0,0,0,0,0,0,0,Larger than 6
239,Etoposide,588.6,161.0,969.0,0.6,42,3,13,5,588.184,588.184,0,1,0,10,10,0,0,0,0,Between -11 and 5
240,Albuterol Sulfate,576.7,228.0,309.0,,39,10,12,10,576.272,576.272,0,3,0,2,0,2,0,0,0,Larger than 6
241,Dexbudesonide,430.5,93.1,862.0,2.5,31,2,6,4,430.236,430.236,0,1,0,9,9,0,0,0,0,Between -11 and 5
242,"2,2-Dimethyl-4-(chloromethyl)-1,3-dioxa-2-silacyclopentane",166.68,18.5,107.0,,9,0,2,1,166.022,166.022,0,1,0,1,0,1,0,0,0,Larger than 6
243,Clopidogrel,321.8,57.8,381.0,3.8,21,0,4,4,321.059,321.059,0,1,0,1,1,0,0,0,0,Between -11 and 5
244,Valsartan,435.5,112.0,608.0,4.4,32,2,6,10,435.227,435.227,0,1,0,1,1,0,0,0,0,Between -11 and 5
245,Tirofiban,440.6,113.0,579.0,1.4,30,3,7,14,440.234,440.234,0,1,0,1,1,0,0,0,0,Between -11 and 5
246,Voriconazole,349.31,76.7,448.0,1.5,25,1,8,5,349.115,349.115,0,1,0,2,2,0,0,0,0,Between -11 and 5
247,N-Phenylethylenediamine,136.19,38.0,77.3,0.6,10,2,2,3,136.1,136.1,0,1,0,0,0,0,0,0,0,Between -11 and 5
248,Oseltamivir phosphate,410.4,168.0,468.0,,27,5,9,8,410.182,410.182,0,2,0,3,3,0,0,0,0,Larger than 6
249,Nicotine,162.23,16.1,147.0,1.2,12,0,2,1,162.116,162.116,0,1,0,1,1,0,0,0,0,Between -11 and 5
250,Lactose monohydrate,360.31,191.0,382.0,,24,9,12,4,360.127,360.127,0,2,0,10,10,0,0,0,0,Larger than 6
251,"2,4-Dioxaspiro[5.5]undec-8-ene, 3-(2-furanyl)-",220.26,31.6,266.0,2.1,16,0,3,1,220.11,220.11,0,1,0,0,0,0,0,0,0,Between -11 and 5
252,Sivelestat,434.5,147.0,731.0,3.0,30,3,8,9,434.115,434.115,0,1,0,0,0,0,0,0,0,Between -11 and 5
253,Tafenoquine,463.5,78.6,597.0,5.4,33,2,9,9,463.208,463.208,0,1,0,1,0,1,0,0,0,Larger than 6
254,IB-Meca,510.3,134.0,589.0,0.9,29,4,8,5,510.051,510.051,0,1,0,4,4,0,0,0,0,Between -11 and 5
255,"(S)-Hexahydropyrrolo[1,2-a]pyrazine-1,4-dione",154.17,49.4,215.0,-0.6,11,1,2,0,154.074,154.074,0,1,0,1,1,0,0,0,0,Between -11 and 5
256,Arbidol,477.4,80.0,546.0,4.4,29,1,5,8,476.077,476.077,0,1,0,0,0,0,0,0,0,Between -11 and 5
257,Tempol,172.24,24.5,159.0,0.9,12,1,2,0,172.134,172.134,0,1,0,0,0,0,0,0,0,Between -11 and 5
258,Atazanavir,704.9,171.0,1110.0,5.6,51,5,9,18,704.39,704.39,0,1,0,4,4,0,0,0,0,Larger than 6
259,Centhaquine,331.5,19.4,404.0,4.5,25,0,3,4,331.205,331.205,0,1,0,0,0,0,0,0,0,Between -11 and 5
260,Regadenoson,390.35,187.0,587.0,-1.5,28,5,10,4,390.14,390.14,0,1,0,4,4,0,0,0,0,Between -11 and 5
261,Teprotide,1101.3,387.0,2330.0,-0.9,79,10,13,24,1100.58,1100.58,0,1,0,10,10,0,0,0,0,Between -11 and 5
262,Azithromycin,749.0,180.0,1150.0,4.0,52,5,14,7,748.509,748.509,0,1,0,18,18,0,0,0,0,Between -11 and 5
263,Posaconazole,700.8,112.0,1170.0,4.6,51,1,11,12,700.33,700.33,0,1,0,4,4,0,0,0,0,Between -11 and 5
264,Cannabidiol,314.5,40.5,414.0,6.5,23,2,2,6,314.225,314.225,0,1,0,2,2,0,0,0,0,Larger than 6
265,"(R)-[2,8-bis(trifluoromethyl)-4-quinolyl]-[(2R)-2-piperidyl]methanol",378.31,45.2,483.0,3.6,26,2,9,2,378.117,378.117,0,1,0,2,2,0,0,0,0,Between -11 and 5
266,Tetramethylol-melamin-dioxy-propylen [German],358.35,170.0,377.0,-0.8,25,6,12,11,358.16,358.16,0,1,0,2,0,2,0,0,0,Between -11 and 5
267,Heme arginate,792.7,260.0,1180.0,,55,8,10,13,792.305,792.305,2,3,0,1,1,0,0,0,0,Larger than 6
268,Zincacetate,185.5,74.6,78.6,,9,2,4,0,183.971,183.971,0,3,0,0,0,0,0,0,0,Larger than 6
269,Montelukast,586.2,95.7,891.0,7.7,41,2,5,12,585.21,585.21,0,1,0,1,1,0,1,1,0,Larger than 6
270,Crocetin,328.4,74.6,608.0,5.4,24,2,4,8,328.167,328.167,0,1,0,0,0,0,7,7,0,Larger than 6
271,Fondaparinux,1508.3,873.0,3450.0,-14.7,91,19,52,30,1506.95,1506.95,0,1,0,25,25,0,0,0,0,Smaller than -10
272,Fluvoxamine,318.33,56.8,327.0,2.6,22,1,7,9,318.156,318.156,0,1,0,0,0,0,1,1,0,Between -11 and 5
273,Naltrexone,341.4,70.0,621.0,1.9,25,2,5,2,341.163,341.163,0,1,0,4,4,0,0,0,0,Between -11 and 5
274,20-Hydroxyecdysone,480.6,138.0,869.0,0.5,34,6,7,5,480.309,480.309,0,1,0,10,10,0,0,0,0,Between -11 and 5
275,Ceftriaxone,554.6,288.0,1110.0,-1.3,36,4,13,8,554.046,554.046,0,1,0,2,2,0,1,1,0,Between -11 and 5
276,"(3E,4S)-4-Hydroxy-3-{2-[(1R,4aS,5R,6R,8aS)-6-hydroxy-5-(hydroxymethyl)-5,8a-dimethyl-2-methylenedecahydronaphthalen-1-yl]ethylidene}dihydrofuran-2(3H)-one",350.4,87.0,597.0,2.2,25,3,5,3,350.209,350.209,0,1,0,6,2,4,1,1,0,Between -11 and 5
277,Artemether and lumefantrine,827.3,69.6,1100.0,,56,1,7,11,825.333,825.333,0,2,0,9,7,2,1,1,0,Larger than 6
278,Isavuconazonium,717.8,188.0,1210.0,4.1,51,2,13,15,717.242,717.242,1,1,0,3,2,1,0,0,0,Between -11 and 5
279,Luminol sodium salt,200.15,84.2,254.0,,14,3,3,0,200.044,200.044,0,2,0,0,0,0,0,0,0,Larger than 6
280,Dapagliflozin,408.9,99.4,472.0,2.3,28,4,6,6,408.134,408.134,0,1,0,5,5,0,0,0,0,Between -11 and 5
281,Zinc Picolinate,309.6,106.0,108.0,,19,0,6,0,307.978,307.978,0,3,0,0,0,0,0,0,0,Larger than 6
282,"Pregna-1,4-diene-3,20-dione,21-(3-carboxy-1-oxopropoxy)-11,17-dihydroxy-6-methyl-, monosodiumsalt, (6a,11b)-",497.5,138.0,981.0,,35,3,8,7,497.215,497.215,0,2,0,8,8,0,0,0,0,Larger than 6
283,Vortioxetine,298.4,40.6,316.0,4.2,21,1,3,3,298.15,298.15,0,1,0,0,0,0,0,0,0,Between -11 and 5
284,Enisamium iodide,354.19,33.0,241.0,,18,1,2,3,354.023,354.023,0,2,0,0,0,0,0,0,0,Larger than 6
285,Cenicriviroc,696.9,105.0,1060.0,7.5,50,1,7,17,696.371,696.371,0,1,0,1,1,0,1,1,0,Larger than 6
286,Apremilast,460.5,128.0,825.0,1.8,32,1,7,8,460.13,460.13,0,1,0,1,1,0,0,0,0,Between -11 and 5
287,Empagliflozin,450.9,109.0,558.0,2.0,31,4,7,6,450.145,450.145,0,1,0,6,6,0,0,0,0,Between -11 and 5
288,N6-ethanimidoyl-D-lysine,187.24,102.0,192.0,-3.1,13,3,4,6,187.132,187.132,0,1,0,1,1,0,0,0,0,Between -11 and 5
289,Emricasan,569.5,151.0,934.0,3.6,40,4,11,11,569.179,569.179,0,1,0,2,2,0,0,0,0,Between -11 and 5
290,Aviptadil Acetate,3344.9,1480.0,7510.0,-13.7,234,51,51,116,3343.74,3342.73,0,1,0,31,0,31,0,0,0,Smaller than -10
291,Quinine sulfate dihydrate,782.9,176.0,538.0,,55,6,14,8,782.356,782.356,0,5,0,8,8,0,0,0,0,Larger than 6
292,Hydrocortisone 21-hemisuccinate sodium salt,485.5,138.0,908.0,,34,3,8,7,485.215,485.215,0,2,0,7,7,0,0,0,0,Larger than 6
293,Liothyronine sodium,672.95,95.6,408.0,,24,2,5,5,672.772,672.772,0,2,0,1,1,0,0,0,0,Larger than 6
294,sodium;8-amino-4-oxo-3H-phthalazin-1-olate,199.14,90.5,269.0,,14,2,4,0,199.036,199.036,0,2,0,0,0,0,0,0,0,Larger than 6
295,Edoxaban tosylate monohydrate,738.3,229.0,1090.0,,49,5,12,6,737.207,737.207,0,3,0,3,3,0,0,0,0,Larger than 6
296,Daclatasvir,738.9,175.0,1190.0,5.1,54,4,8,13,738.385,738.385,0,1,0,4,4,0,0,0,0,Larger than 6
297,"2-[(4S)-4-amino-5,5-dihydroxyhexyl]guanidine",190.24,131.0,177.0,-2.5,13,5,4,5,190.143,190.143,0,1,0,1,1,0,0,0,0,Between -11 and 5
298,Fosmanogepix,468.4,148.0,644.0,1.6,33,2,9,9,468.12,468.12,0,1,0,0,0,0,0,0,0,Between -11 and 5
299,Larazotide acetate,785.9,339.0,1320.0,,55,10,13,21,785.428,785.428,0,2,0,5,5,0,0,0,0,Larger than 6
300,Thymosin,3051.3,1370.0,6930.0,-22.0,212,47,57,109,3050.5,3049.5,0,1,0,30,30,0,0,0,0,Smaller than -10
301,Sofosbuvir,529.5,153.0,913.0,1.0,36,3,11,11,529.163,529.163,0,1,0,6,6,0,0,0,0,Between -11 and 5
302,Razuprotafib,586.7,212.0,906.0,3.7,39,4,10,12,586.101,586.101,0,1,0,2,2,0,0,0,0,Between -11 and 5
303,Nalpha-[(4-Methylpiperazin-1-Yl)carbonyl]-N-[(3s)-1-Phenyl-5-(Phenylsulfonyl)pentan-3-Yl]-L-Phenylalaninamide,576.8,107.0,897.0,4.4,41,2,5,12,576.277,576.277,0,1,0,2,2,0,0,0,0,Between -11 and 5
304,Ascorbic Acid,176.12,107.0,232.0,-1.6,12,4,6,2,176.032,176.032,0,1,0,2,2,0,0,0,0,Between -11 and 5
305,Doxycycline,444.4,182.0,956.0,-0.7,32,6,9,2,444.153,444.153,0,1,0,6,6,0,0,0,0,Between -11 and 5
306,Rabeximod,409.9,63.0,590.0,3.9,29,1,4,5,409.167,409.167,0,1,0,0,0,0,0,0,0,Between -11 and 5
307,Kaolin,258.16,98.0,167.0,,13,2,9,4,257.902,257.902,0,3,0,0,0,0,0,0,0,Larger than 6
308,H-Ile-OH.H-Thr-OH.H-Leu-OH.H-Val-OH.H-Met-OH.H-Phe-OH.H-Trp-OH.H-Lys-OH,1163.4,594.0,988.0,,80,19,27,25,1162.65,1162.65,0,8,0,10,10,0,0,0,0,Larger than 6
309,Semaglutide,4114.0,1650.0,9590.0,-5.8,291,57,63,151,4112.12,4111.12,0,1,0,30,30,0,0,0,0,Between -11 and 5
310,Isuzinaxib hydrochloride,315.8,45.2,412.0,,22,2,3,4,315.114,315.114,0,2,0,0,0,0,0,0,0,Larger than 6
311,CID 66726979,457.5,147.0,731.0,,31,3,8,9,457.105,457.105,0,2,0,0,0,0,0,0,0,Larger than 6
312,Desidustat,332.31,116.0,583.0,1.9,24,3,6,6,332.101,332.101,0,1,0,0,0,0,0,0,0,Between -11 and 5
313,"7-[[2-Ethoxyimino-2-[5-(phosphonoamino)-1,2,4-thiadiazol-3-yl]acetyl]amino]-3-[[4-(1-methylpyridin-1-ium-4-yl)-1,3-thiazol-2-yl]sulfanyl]-8-oxo-5-thia-1-azabicyclo[4.2.0]oct-2-ene-2-carboxylic acid;acetate",744.7,368.0,1240.0,,47,5,19,11,744.031,744.031,0,2,0,2,0,2,1,0,1,Larger than 6
314,1-(4-(((6-Amino-5-(4-phenoxyphenyl)pyrimidin-4-yl)amino)methyl)-4-fluoropiperidin-1-yl)prop-2-en-1-one,447.5,93.4,643.0,4.1,33,2,7,7,447.207,447.207,0,1,0,0,0,0,0,0,0,Between -11 and 5
315,Ceftriaxone disodium hemiheptahydrate,600.6,288.0,1110.0,,38,4,13,8,600.026,600.026,0,3,0,2,2,0,1,1,0,Larger than 6
316,Descovy,723.7,257.0,1050.0,,49,4,15,14,723.236,723.236,0,2,0,5,5,0,0,0,0,Larger than 6
317,Zimlovisertib,361.4,104.0,535.0,2.0,26,2,6,6,361.144,361.144,0,1,0,3,3,0,0,0,0,Between -11 and 5
318,CID 131673872,202.17,84.2,254.0,,14,3,3,0,202.059,202.059,0,3,0,0,0,0,0,0,0,Larger than 6
319,"Sodium;5-amino-2,3-dihydrophthalazine-1,4-dione;hydride",201.16,84.2,254.0,,14,3,4,0,201.051,201.051,0,3,0,0,0,0,0,0,0,Larger than 6
320,Zilucoplan,3562.0,1070.0,6980.0,4.8,251,28,57,142,3560.97,3559.97,0,1,0,16,16,0,0,0,0,Between -11 and 5
321,Defibrotide,444.4,137.0,773.0,1.8,31,4,7,5,444.12,444.12,0,1,0,0,0,0,0,0,0,Between -11 and 5
322,"7-[(3s,4r)-4-(3-Chlorophenyl)carbonylpyrrolidin-3-Yl]-3h-Quinazolin-4-One",353.8,70.6,567.0,2.1,25,2,4,3,353.093,353.093,0,1,0,2,2,0,0,0,0,Between -11 and 5
323,Unii-7kyp9tkt70,879.6,484.0,1210.0,,58,15,24,13,879.205,879.205,0,4,0,8,0,8,0,0,0,Larger than 6
324,"4-acetamidobenzoic acid;9-[(2R,3R,4R,5R)-3,4-dihydroxy-5-(hydroxymethyl)oxolan-2-yl]-1H-purin-6-one;1-(dimethylamino)propan-2-ol",1115.2,399.0,658.0,,79,13,22,14,1114.55,1114.55,0,7,0,7,4,3,0,0,0,Larger than 6
325,"[(1S,4R,6S)-4-(cyclopropylsulfonylcarbamoyl)-14-[(2-methylpropan-2-yl)oxycarbonylamino]-2,15-dioxo-3,16-diazatricyclo[14.3.0.04,6]nonadec-7-en-18-yl] 4-fluoro-1,3-dihydroisoindole-2-carboxylate",731.8,189.0,1530.0,3.3,51,3,10,8,731.3,731.3,0,1,0,5,3,2,1,0,1,Between -11 and 5
326,example 13 [US20210284598A1],257.279,40.5,287.0,2.7,18,1,4,4,257.123,257.123,0,1,0,0,0,0,0,0,0,Between -11 and 5
327,Stannous protoporphyrin,679.4,102.0,1010.0,,43,2,8,8,680.145,680.145,0,2,0,0,0,0,0,0,0,Larger than 6
328,"3,6-Di-O-acetyl-2-deoxy-d-glucopyranose",248.23,110.0,276.0,-1.7,17,2,7,9,248.09,248.09,0,1,0,3,3,0,0,0,0,Between -11 and 5
329,X6 hydrobromide [PMID: 34584244],502.5,168.0,887.0,2.3,36,1,11,11,502.229,502.229,0,1,1,4,4,0,0,0,0,Between -11 and 5
330,Enoxaparin,1134.9,652.0,2410.0,-10.8,70,15,38,21,1134.01,1134.01,0,1,0,20,0,20,0,0,0,Smaller than -10
331,Amantadine,151.25,26.0,144.0,2.4,11,1,1,0,151.136,151.136,0,1,0,0,0,0,0,0,0,Between -11 and 5
332,Amlodipine,408.9,99.9,647.0,3.0,28,2,7,10,408.145,408.145,0,1,0,1,0,1,0,0,0,Between -11 and 5
333,Bicalutamide,430.4,116.0,750.0,2.3,29,2,9,5,430.061,430.061,0,1,0,1,0,1,0,0,0,Between -11 and 5
334,Candesartan cilexetil,610.7,143.0,962.0,7.0,45,1,10,13,610.254,610.254,0,1,0,1,0,1,0,0,0,Larger than 6
335,Formoterol,344.4,90.8,388.0,1.8,25,4,5,8,344.174,344.174,0,1,0,2,0,2,0,0,0,Between -11 and 5
336,Hydroxychloroquine,335.9,48.4,331.0,3.6,23,2,4,9,335.176,335.176,0,1,0,1,0,1,0,0,0,Between -11 and 5
337,Ibudilast,230.31,34.4,288.0,3.0,17,0,2,3,230.142,230.142,0,1,0,0,0,0,0,0,0,Between -11 and 5
338,Lidocaine,234.34,32.299,228.0,2.3,17,1,2,5,234.173,234.173,0,1,0,0,0,0,0,0,0,Between -11 and 5
339,Modafinil,273.4,79.4,302.0,1.7,19,1,3,5,273.082,273.082,0,1,0,1,0,1,0,0,0,Between -11 and 5
340,Omeprazole,345.4,96.3,453.0,2.2,24,1,6,5,345.115,345.115,0,1,0,1,0,1,0,0,0,Between -11 and 5
341,Pentoxifylline,278.31,75.5,426.0,0.3,20,0,4,5,278.138,278.138,0,1,0,0,0,0,0,0,0,Between -11 and 5
342,Arginine,174.2,128.0,176.0,-4.2,12,4,4,5,174.112,174.112,0,1,0,1,1,0,0,0,0,Between -11 and 5
343,"4,4'-Diphenylmethane diisocyanate",250.25,58.9,332.0,5.4,19,0,4,4,250.074,250.074,0,1,0,0,0,0,0,0,0,Larger than 6
344,Carvacrol,150.22,20.2,120.0,3.1,11,1,1,1,150.104,150.104,0,1,0,0,0,0,0,0,0,Between -11 and 5
345,Silver,107.868,0.0,0.0,,1,0,0,0,106.905,106.905,0,1,0,0,0,0,0,0,0,Larger than 6
346,Phorbol 12-myristate 13-acetate,616.8,130.0,1150.0,6.5,44,3,8,17,616.398,616.398,0,1,0,8,8,0,0,0,0,Larger than 6
347,(-)-Mefloquine,378.31,45.2,483.0,3.6,26,2,9,2,378.117,378.117,0,1,0,2,2,0,0,0,0,Between -11 and 5
348,Atorvastatin,558.6,112.0,822.0,5.0,41,4,6,12,558.253,558.253,0,1,0,2,2,0,0,0,0,Between -11 and 5
349,Propranolol Hydrochloride,295.8,41.5,257.0,,20,3,3,6,295.134,295.134,0,2,0,1,0,1,0,0,0,Larger than 6
350,Oseltamivir,312.4,90.6,418.0,1.1,22,2,5,8,312.205,312.205,0,1,0,3,3,0,0,0,0,Between -11 and 5
351,"2,4,6-Trinitro-m-xylene",241.16,138.0,317.0,2.0,17,0,6,0,241.033,241.033,0,1,0,0,0,0,0,0,0,Between -11 and 5
352,Argatroban,508.6,189.0,887.0,1.3,35,5,8,9,508.247,508.247,0,1,0,4,3,1,0,0,0,Between -11 and 5
353,Lopinavir,628.8,120.0,940.0,5.9,46,4,5,15,628.362,628.362,0,1,0,4,4,0,0,0,0,Larger than 6
354,Allopregnanolone,318.5,37.3,500.0,4.9,23,1,2,1,318.256,318.256,0,1,0,8,8,0,0,0,0,Between -11 and 5
355,Fingolimod,307.5,66.5,258.0,4.2,22,3,3,12,307.251,307.251,0,1,0,0,0,0,0,0,0,Between -11 and 5
356,Imatinib Mesylate,589.7,149.0,799.0,,42,3,10,7,589.247,589.247,0,2,0,0,0,0,0,0,0,Larger than 6
357,Meldonium,146.19,52.2,112.0,-2.1,10,1,3,3,146.106,146.106,0,1,0,0,0,0,0,0,0,Between -11 and 5
358,Ramatroban,416.5,96.8,689.0,2.9,29,2,6,6,416.121,416.121,0,1,0,1,1,0,0,0,0,Between -11 and 5
359,Ivabradine,468.6,60.5,663.0,2.4,34,0,6,10,468.262,468.262,0,1,0,1,1,0,0,0,0,Between -11 and 5
360,Moxifloxacin,401.4,82.1,727.0,0.6,29,2,8,4,401.175,401.175,0,1,0,2,2,0,0,0,0,Between -11 and 5
361,Varespladib,380.4,112.0,589.0,2.8,28,2,5,8,380.137,380.137,0,1,0,0,0,0,0,0,0,Between -11 and 5
362,Naproxen,230.26,46.5,277.0,3.3,17,1,3,3,230.094,230.094,0,1,0,1,1,0,0,0,0,Between -11 and 5
363,Dabigatran,471.5,150.0,757.0,1.7,35,4,7,9,471.202,471.202,0,1,0,0,0,0,0,0,0,Between -11 and 5
364,Senicapoc,323.3,43.1,397.0,4.1,24,1,3,4,323.112,323.112,0,1,0,0,0,0,0,0,0,Between -11 and 5
365,Povidone iodine,364.95,20.3,120.0,,10,0,1,1,364.877,364.877,0,2,0,0,0,0,0,0,0,Larger than 6
366,beta-L-Arabinose,150.13,90.2,117.0,-2.5,10,4,5,0,150.053,150.053,0,1,0,4,4,0,0,0,0,Between -11 and 5
367,Quinidine,324.4,45.6,457.0,2.9,24,1,4,4,324.184,324.184,0,1,0,4,4,0,0,0,0,Between -11 and 5
368,Pectin,194.14,127.0,205.0,-2.3,13,5,7,1,194.043,194.043,0,1,0,5,5,0,0,0,0,Between -11 and 5
369,Fluticasone Propionate,500.6,106.0,984.0,4.0,34,1,9,6,500.184,500.184,0,1,0,9,9,0,0,0,0,Between -11 and 5
370,Decitabine,228.21,121.0,356.0,-1.2,16,3,4,2,228.086,228.086,0,1,0,3,3,0,0,0,0,Between -11 and 5
371,Bucillamine,223.3,68.4,218.0,0.4,13,4,5,4,223.034,223.034,0,1,0,1,1,0,0,0,0,Between -11 and 5
372,Canrenoic acid,358.5,74.6,707.0,1.9,26,2,4,3,358.214,358.214,0,1,0,6,6,0,0,0,0,Between -11 and 5
373,"(S)-[2,8-bis(trifluoromethyl)quinolin-4-yl]-[(2S)-piperidin-2-yl]methanol",378.31,45.2,483.0,3.6,26,2,9,2,378.117,378.117,0,1,0,2,2,0,0,0,0,Between -11 and 5
374,"(-)-(S)-9-Fluoro-2,3-dihydro-3-methyl-10-(4-methyl-1-piperazinyl)-7-oxo-7H-pyrido(1,2,3-de)-1,4-benzoxazine-6-carboxylic acid, hemihydrate",740.7,148.0,634.0,,53,3,17,4,740.298,740.298,0,3,0,2,2,0,0,0,0,Larger than 6
375,"(S,S)-Formoterol",344.4,90.8,388.0,1.8,25,4,5,8,344.174,344.174,0,1,0,2,2,0,0,0,0,Between -11 and 5
376,Glatiramer acetate,623.7,374.0,519.0,,43,12,18,13,623.301,623.301,0,5,0,4,4,0,0,0,0,Larger than 6
377,Isoquercetin,464.4,207.0,758.0,0.4,33,8,12,4,464.095,464.095,0,1,0,5,5,0,0,0,0,Between -11 and 5
378,Pitavastatin,421.5,90.6,631.0,3.5,31,3,6,8,421.169,421.169,0,1,0,2,2,0,1,1,0,Between -11 and 5
379,Deoxy-methyl-arginine,172.23,108.0,174.0,-1.8,12,3,3,5,172.132,172.132,0,1,0,1,1,0,0,0,0,Between -11 and 5
380,Dexmedetomidine,200.28,28.7,205.0,3.1,15,1,1,2,200.131,200.131,0,1,0,1,1,0,0,0,0,Between -11 and 5
381,Nafamostat mesylate,539.6,266.0,645.0,,36,6,10,5,539.114,539.114,0,3,0,0,0,0,0,0,0,Larger than 6
382,Bromhexine Hydrochloride,412.59,29.3,256.0,,19,2,2,3,411.974,409.976,0,2,0,0,0,0,0,0,0,Larger than 6
383,Tenofovir Disoproxil Fumarate,635.5,260.0,817.0,,43,3,18,19,635.184,635.184,0,2,0,1,1,0,1,1,0,Larger than 6
384,Eritoran,1313.7,294.0,1900.0,15.4,89,7,19,59,1312.84,1312.84,0,1,0,11,11,0,1,1,0,Larger than 6
385,Icatibant acetate,1364.6,589.0,2750.0,,96,16,20,30,1363.68,1363.68,0,2,0,12,12,0,0,0,0,Larger than 6
386,Dalcetrapib,389.6,71.5,481.0,7.1,27,1,3,9,389.239,389.239,0,1,0,0,0,0,0,0,0,Larger than 6
387,"7-[[(2Z)-2-ethoxyimino-2-[5-(phosphonoamino)-1,2,4-thiadiazol-3-yl]acetyl]amino]-3-[[4-(1-methylpyridin-1-ium-4-yl)-1,3-thiazol-2-yl]sulfanyl]-8-oxo-5-thia-1-azabicyclo[4.2.0]oct-2-ene-2-carboxylic acid",685.7,328.0,1220.0,1.6,43,5,17,11,685.018,685.018,1,1,0,2,0,2,1,1,0,Between -11 and 5
388,Icosapent ethyl,330.5,26.3,425.0,6.3,24,0,2,15,330.256,330.256,0,1,0,0,0,0,5,5,0,Larger than 6
389,"1-Piperazinecarboxamide, 4-methyl-N-((1S)-2-oxo-2-(((1S)-1-(2-phenylethyl)-3-(phenylsulfonyl)-2-propenyl)amino)-1-(phenylmethyl)ethyl)-",574.7,107.0,939.0,4.1,41,2,5,11,574.261,574.261,0,1,0,2,2,0,1,1,0,Between -11 and 5
390,Remimazolam,439.3,69.4,601.0,3.4,28,0,5,5,438.069,438.069,0,1,0,1,1,0,0,0,0,Between -11 and 5
391,Azithromycin Monohydrate,767.0,181.0,1150.0,,53,6,15,7,766.519,766.519,0,2,0,18,18,0,0,0,0,Larger than 6
392,Danoprevir,731.8,189.0,1530.0,3.3,51,3,10,8,731.3,731.3,0,1,0,5,5,0,1,1,0,Between -11 and 5
393,Mitoquinone,583.7,52.6,886.0,9.4,42,0,4,16,583.298,583.298,1,1,0,0,0,0,0,0,0,Larger than 6
394,Belnacasan,509.0,140.0,818.0,2.3,35,3,7,8,508.209,508.209,0,1,0,4,4,0,0,0,0,Between -11 and 5
395,Apilimod mesylate,610.7,210.0,637.0,,41,3,14,8,610.188,610.188,0,3,0,0,0,0,1,1,0,Larger than 6
396,Losmapimod,383.5,71.1,573.0,3.8,28,2,4,6,383.201,383.201,0,1,0,0,0,0,0,0,0,Between -11 and 5
397,Emtricitabine and tenofovir disoproxil fumarate,882.8,374.0,1190.0,,59,5,23,21,882.227,882.227,0,3,0,3,3,0,1,1,0,Larger than 6
398,Dapansutrile,133.17,66.3,190.0,-0.7,8,0,3,2,133.02,133.02,0,1,0,0,0,0,0,0,0,Between -11 and 5
399,"(S)-[2,8-bis(trifluoromethyl)quinolin-4-yl]-piperidin-2-ylmethanol",378.31,45.2,483.0,3.6,26,2,9,2,378.117,378.117,0,1,0,2,1,1,0,0,0,Between -11 and 5
400,Tridecactide,1623.8,659.0,3240.0,-6.2,115,23,24,50,1622.77,1622.77,0,1,0,12,12,0,0,0,0,Between -11 and 5
401,Aviptadil,3326.8,1470.0,7580.0,-15.9,234,51,51,115,3325.74,3324.74,0,1,0,31,31,0,0,0,0,Smaller than -10
402,Pamapimod,406.4,108.0,591.0,2.4,29,3,9,8,406.145,406.145,0,1,0,0,0,0,0,0,0,Between -11 and 5
403,Sodium pyruvate,110.04,57.2,88.2,,7,0,3,1,109.998,109.998,0,2,0,0,0,0,0,0,0,Larger than 6
404,Brequinar sodium,397.3,53.0,557.0,,29,0,5,3,397.089,397.089,0,2,0,0,0,0,0,0,0,Larger than 6
405,Montelukast Sodium,608.2,98.6,898.0,,42,1,5,12,607.192,607.192,0,2,0,1,1,0,1,1,0,Larger than 6
406,Sivelestat sodium anhydrous,456.4,150.0,738.0,,31,2,8,9,456.097,456.097,0,2,0,0,0,0,0,0,0,Larger than 6
407,Methylprednisolone sodium succinate,496.5,141.0,988.0,,35,2,8,7,496.207,496.207,0,2,0,8,8,0,0,0,0,Larger than 6
408,Bromelains,248.25,81.6,344.0,,17,1,3,4,248.114,248.114,0,2,0,2,0,2,0,0,0,Larger than 6
409,Fostamatinib disodium,732.5,198.0,893.0,,48,8,21,9,732.176,732.176,0,9,0,0,0,0,0,0,0,Larger than 6
410,Phosphate-Buffered Saline,411.04,164.0,96.3,,17,3,10,0,409.765,409.765,0,9,0,0,0,0,0,0,0,Larger than 6
411,4'-C-Azido-2'-deoxy-2'-fluoro-b-D-arabinocytidine,286.22,123.0,533.0,-0.8,20,3,7,3,286.083,286.083,0,1,0,4,0,4,1,0,1,Between -11 and 5
412,5-Chloro-2-([(2-([3-(furan-2-yl)phenyl]amino)-2-oxoethoxy)acetyl]amino)benzoic acid,428.8,118.0,618.0,3.3,30,3,6,8,428.078,428.078,0,1,0,0,0,0,0,0,0,Between -11 and 5
413,"[(2R,3S,4R)-4-acetyloxy-3,6-dihydroxyoxan-2-yl]methyl acetate",248.23,102.0,290.0,-1.0,17,2,7,5,248.09,248.09,0,1,0,4,3,1,0,0,0,Between -11 and 5
414,Ketone Ester,176.21,66.8,135.0,-0.1,12,2,4,6,176.105,176.105,0,1,0,2,2,0,0,0,0,Between -11 and 5
415,CID 45114162,520.6,152.0,761.0,,34,2,10,9,520.123,520.123,1,2,0,2,2,0,1,1,0,Larger than 6
416,Alvelestat,545.5,123.0,1100.0,2.5,38,1,9,6,545.134,545.134,0,1,0,0,0,0,0,0,0,Between -11 and 5
417,Emvododstat,467.3,54.6,651.0,6.3,32,1,3,4,466.085,466.085,0,1,0,1,1,0,0,0,0,Larger than 6
418,8-chloro-N-[4-(trifluoromethoxy)phenyl]quinolin-2-amine,338.71,34.2,388.0,5.9,23,1,6,3,338.043,338.043,0,1,0,0,0,0,0,0,0,Larger than 6
419,Riamilovir sodium dihydrate,286.2,152.0,262.0,,18,2,10,1,286.01,286.01,0,4,0,0,0,0,0,0,0,Larger than 6
420,Doxycycline hyclate,545.0,203.0,958.0,,37,9,11,2,544.182,544.182,0,4,0,6,6,0,0,0,0,Larger than 6
421,Budesonide and formoterol fumarate dihydrate,774.9,184.0,1250.0,,56,6,11,12,774.409,774.409,0,2,0,11,10,1,0,0,0,Larger than 6
422,Vafidemstat,336.4,86.2,410.0,2.2,25,2,6,7,336.159,336.159,0,1,0,2,2,0,0,0,0,Between -11 and 5
423,Ledipasvir,889.0,175.0,1820.0,7.4,65,4,10,12,888.413,888.413,0,1,0,6,6,0,0,0,0,Larger than 6
424,Telacebec,557.0,58.9,796.0,7.9,39,1,7,7,556.185,556.185,0,1,0,0,0,0,0,0,0,Larger than 6
425,Acebilustat,481.5,79.0,728.0,2.2,36,1,7,8,481.2,481.2,0,1,0,2,2,0,0,0,0,Between -11 and 5
426,Quinidine monohydrate,342.4,46.6,457.0,,25,2,5,4,342.194,342.194,0,2,0,4,4,0,0,0,0,Larger than 6
427,Dexamethasone phosphate disodium,518.4,141.0,973.0,,34,4,9,4,518.146,518.146,0,3,0,8,8,0,0,0,0,Larger than 6
428,Montmorillonite,360.31,141.0,18.3,,18,1,12,0,359.825,359.825,0,10,0,0,0,0,0,0,0,Larger than 6
429,methyl N-[(2R)-2-[2-[5-[4-[4-[2-[1-[(2S)-2-(methoxycarbonylamino)-3-methyl-butanoyl]pyrrolidin-2-yl]-1H-imidazol-5-yl]phenyl]phenyl]-1H-imidazol-2-yl]pyrrolidine-1-carbonyl]-3-methyl-butyl]carbamate,752.9,175.0,1300.0,5.0,55,4,8,14,752.401,752.401,0,1,0,4,2,2,0,0,0,Between -11 and 5
430,Nadide sodium,685.4,324.0,1110.0,,45,6,18,11,685.091,685.091,0,2,0,8,8,0,0,0,0,Larger than 6
431,"(6R,7R)-7-[[(2E)-2-ethoxyimino-2-[5-(phosphonoamino)-1,2,4-thiadiazol-3-yl]acetyl]amino]-3-[[4-(1-methylpyridin-1-ium-4-yl)-1,3-thiazol-2-yl]sulfanyl]-8-oxo-5-thia-1-azabicyclo[4.2.0]oct-2-ene-2-carboxylate",684.7,330.0,1210.0,2.3,43,4,17,10,684.01,684.01,0,1,0,2,2,0,1,1,0,Between -11 and 5
432,"3-Deoxy-3-[4-(3-Fluorophenyl)-1h-1,2,3-Triazol-1-Yl]-Beta-D-Galactopyranosyl 3-Deoxy-3-[4-(3-Fluorophenyl)-1h-1,2,3-Triazol-1-Yl]-1-Thio-Beta-D-Galactopyranoside",648.6,227.0,903.0,0.2,45,6,15,8,648.181,648.181,0,1,0,10,10,0,0,0,0,Between -11 and 5
433,"N-[(2S,3S,4R)-1-(alpha-D-galactopyranosyloxy)-3,4-dihydroxyoctadecan-2-yl]undecanamide",647.9,169.0,700.0,7.9,45,7,9,29,647.497,647.497,0,1,0,8,8,0,0,0,0,Larger than 6
434,Brensocatib,420.5,104.0,699.0,2.0,31,2,6,5,420.18,420.18,0,1,0,2,2,0,0,0,0,Between -11 and 5
435,Ezurpimtrostat,423.0,40.2,520.0,5.9,30,2,4,6,422.224,422.224,0,1,0,0,0,0,0,0,0,Larger than 6
436,Dexamethasone 21-phosphate disodium salt,548.5,141.0,973.0,,36,4,9,4,548.193,548.193,0,4,0,8,8,0,0,0,0,Larger than 6
437,Enpatoran,320.31,65.9,472.0,2.7,23,1,7,1,320.125,320.125,0,1,0,2,2,0,0,0,0,Between -11 and 5
438,H-D-Lys-D-Leu-D-Leu-D-Leu-D-Leu-D-Lys-D-Leu-D-Leu-D-Leu-D-Leu-D-Lys-D-Leu-D-Leu-D-Leu-D-Leu-D-Lys-D-Leu-D-Leu-D-Leu-D-Leu-D-Lys-OH,2469.4,775.0,4880.0,11.9,174,27,28,93,2468.83,2467.83,0,1,0,21,21,0,0,0,0,Larger than 6
439,CID 131844884,1738.2,873.0,3450.0,,101,19,52,30,1736.85,1736.85,0,11,0,25,25,0,0,0,0,Larger than 6
440,Folic Acid,441.4,209.0,767.0,-1.1,32,6,10,9,441.14,441.14,0,1,0,1,1,0,0,0,0,Between -11 and 5
441,Inosine pranobex,1115.2,399.0,658.0,,79,13,22,14,1114.55,1114.55,0,7,0,7,4,3,0,0,0,Larger than 6
442,Eltrombopag,442.5,115.0,812.0,5.4,33,3,7,5,442.164,442.164,0,1,0,0,0,0,0,0,0,Larger than 6
443,Rintatolimod,995.6,517.0,1600.0,,65,15,27,12,995.135,995.135,0,3,0,12,12,0,0,0,0,Larger than 6
444,Normosang,792.7,260.0,1180.0,,55,8,10,13,792.305,792.305,2,3,0,1,1,0,0,0,0,Larger than 6
445,Artecom,1294.4,485.0,1380.0,,84,15,31,12,1293.31,1293.31,0,6,0,8,8,0,0,0,0,Larger than 6
446,Imunovir; Delimmun; Groprinosin;Inosine pranobex,1115.2,399.0,658.0,,79,13,22,14,1114.55,1114.55,0,7,0,7,0,7,0,0,0,Larger than 6
447,"1-(4-fluorobenzene-6-id-1-yl)-N-phenylmethanimine;iridium(3+);1,2,3,4,5-pentamethylcyclopenta-1,3-diene;chloride",561.1,12.4,601.0,,27,0,5,1,561.121,561.121,0,4,0,0,0,0,0,0,0,Larger than 6
448,EIDD-2801,329.31,141.0,534.0,-0.8,23,4,7,6,329.122,329.122,0,1,0,4,4,0,0,0,0,Between -11 and 5
449,CID 154701548,287.21,143.0,435.0,,18,3,8,1,287.017,287.017,0,4,0,0,0,0,0,0,0,Larger than 6
450,"alpha-D-Glucopyranoside, methyl O-2-deoxy-6-O-sulfo-2-(sulfoamino)-alpha-D-glucopyranosyl-(1-->4)-O-beta-D-glucopyranuronosyl-(1-->4)-O-2-deoxy-3,6-di-O-sulfo-2-(sulfoamino)-alpha-D-glucopyranosyl-(1-->4)-O-2-O-sulfo-alpha-L-idopyranuronosyl-(1-->4)-2-deoxy-2-(sulfoamino)-, 6-(hydrogen sulfate), sodium salt (1:10)",1531.3,873.0,3450.0,,92,19,52,30,1529.94,1529.94,0,2,0,25,25,0,0,0,0,Larger than 6
451,"propan-2-yl (2S)-2-[[[(2R,3R,4R,5R)-5-[2-amino-6-(methylamino)purin-9-yl]-4-fluoro-3-hydroxy-4-methyloxolan-2-yl]methoxy-phenoxyphosphoryl]amino]propanoate;sulfuric acid",679.6,268.0,1000.0,,45,6,18,12,679.184,679.184,0,2,0,6,5,1,0,0,0,Larger than 6
452,Ensitrelvir,531.9,114.0,919.0,2.5,37,1,8,6,531.115,531.115,0,1,0,0,0,0,0,0,0,Between -11 and 5
453,"3,4-Methylenedioxymethamphetamine",193.24,30.5,186.0,2.2,14,1,3,3,193.11,193.11,0,1,0,1,0,1,0,0,0,Between -11 and 5
454,Acetaminophen,151.16,49.3,139.0,0.5,11,2,2,1,151.063,151.063,0,1,0,0,0,0,0,0,0,Between -11 and 5
455,Amiodarone,645.3,42.7,547.0,7.6,31,0,4,11,645.024,645.024,0,1,0,0,0,0,0,0,0,Larger than 6
456,Verapamil,454.6,64.0,606.0,3.8,33,0,6,13,454.283,454.283,0,1,0,1,0,1,0,0,0,Between -11 and 5
457,Candesartan,440.5,119.0,660.0,4.1,33,2,7,7,440.16,440.16,0,1,0,0,0,0,0,0,0,Between -11 and 5
458,Chlordiazepoxide,299.75,48.2,580.0,2.4,21,1,3,1,299.083,299.083,0,1,0,0,0,0,0,0,0,Between -11 and 5
459,Chloroquine,319.9,28.2,309.0,4.6,22,1,3,8,319.182,319.182,0,1,0,1,0,1,0,0,0,Between -11 and 5
460,Deferoxamine,560.7,206.0,739.0,-2.1,39,6,9,23,560.353,560.353,0,1,0,0,0,0,0,0,0,Between -11 and 5
461,Famotidine,337.5,238.0,469.0,-0.6,20,4,8,7,337.045,337.045,0,1,0,0,0,0,1,0,1,Between -11 and 5
462,Fenofibrate,360.8,52.6,458.0,5.2,25,0,4,7,360.113,360.113,0,1,0,0,0,0,0,0,0,Larger than 6
463,Ketamine,237.72,29.1,269.0,2.2,16,1,2,2,237.092,237.092,0,1,0,1,0,1,0,0,0,Between -11 and 5
464,Lansoprazole,369.4,87.1,480.0,2.8,25,1,8,5,369.076,369.076,0,1,0,1,0,1,0,0,0,Between -11 and 5
465,Metformin,129.16,91.5,132.0,-1.3,9,3,1,2,129.101,129.101,0,1,0,0,0,0,0,0,0,Between -11 and 5
466,Nafamostat,347.4,141.0,552.0,2.0,26,4,4,5,347.138,347.138,0,1,0,0,0,0,0,0,0,Between -11 and 5
467,Palmitoylethanolamide,299.5,49.3,219.0,6.2,21,2,2,16,299.282,299.282,0,1,0,0,0,0,0,0,0,Larger than 6
468,Pioglitazone,356.4,93.6,466.0,3.8,25,1,5,7,356.119,356.119,0,1,0,1,0,1,0,0,0,Between -11 and 5
469,Propofol,178.27,20.2,135.0,3.8,13,1,1,2,178.136,178.136,0,1,0,0,0,0,0,0,0,Between -11 and 5
470,Sevoflurane,200.05,9.2,121.0,2.8,12,0,8,2,200.007,200.007,0,1,0,0,0,0,0,0,0,Between -11 and 5
471,Tranexamic acid,157.21,63.3,139.0,-2.0,11,2,3,2,157.11,157.11,0,1,0,0,0,0,0,0,0,Between -11 and 5
472,Prednisolone,360.4,94.8,724.0,1.6,26,3,5,2,360.194,360.194,0,1,0,7,7,0,0,0,0,Between -11 and 5
473,Levothyroxine,776.87,92.8,420.0,2.4,24,3,5,5,776.687,776.687,0,1,0,1,1,0,0,0,0,Between -11 and 5
474,Mannitol,182.17,121.0,105.0,-3.1,12,6,6,5,182.079,182.079,0,1,0,4,4,0,0,0,0,Between -11 and 5
475,Dexamethasone phosphate,472.4,141.0,973.0,0.1,32,4,9,4,472.166,472.166,0,1,0,8,8,0,0,0,0,Between -11 and 5
476,Citrulline,175.19,118.0,171.0,-4.3,12,4,4,5,175.096,175.096,0,1,0,1,1,0,0,0,0,Between -11 and 5
477,Hesperidin,610.6,234.0,940.0,-1.1,43,8,15,7,610.19,610.19,0,1,0,11,11,0,0,0,0,Between -11 and 5
478,Zinc Citrate,574.3,281.0,211.0,,29,2,14,4,571.791,569.794,0,5,0,0,0,0,0,0,0,Larger than 6
479,Zinc Acetate,183.5,80.3,25.5,,9,0,4,0,181.956,181.956,0,3,0,0,0,0,0,0,0,Larger than 6
480,Canrenone,340.5,43.4,719.0,2.7,25,0,3,0,340.204,340.204,0,1,0,6,6,0,0,0,0,Between -11 and 5
481,Hydrocortisone hemisuccinate,462.5,138.0,908.0,1.8,33,3,8,7,462.225,462.225,0,1,0,7,7,0,0,0,0,Between -11 and 5
482,Methylprednisolone hemisuccinate,474.5,138.0,981.0,2.2,34,3,8,7,474.225,474.225,0,1,0,8,8,0,0,0,0,Between -11 and 5
483,Azo rubin S,458.5,170.0,880.0,3.3,31,3,9,4,458.024,458.024,0,1,0,0,0,0,0,0,0,Between -11 and 5
484,"Zinc, bis(D-gluconato-kappaO1,kappaO2)-, (T-4)-",457.7,277.0,385.0,,27,12,14,10,456.046,456.046,0,3,0,8,8,0,0,0,0,Larger than 6
485,2-acetamido-2-deoxy-beta-D-glucopyranose,221.21,119.0,235.0,-1.7,15,5,6,2,221.09,221.09,0,1,0,5,5,0,0,0,0,Between -11 and 5
486,Zinc Sulfate,161.4,88.6,62.2,,6,0,4,0,159.881,159.881,0,2,0,0,0,0,0,0,0,Larger than 6
487,Sodium Thiosulfate,158.11,104.0,82.6,,7,0,4,0,157.908,157.908,0,3,0,0,0,0,0,0,0,Larger than 6
488,Diacerein,368.3,124.0,683.0,1.9,27,1,8,5,368.053,368.053,0,1,0,0,0,0,0,0,0,Between -11 and 5
489,Levamisole,204.29,40.9,246.0,1.8,14,0,2,1,204.072,204.072,0,1,0,1,1,0,0,0,0,Between -11 and 5
490,Cromolyn sodium,512.299,172.0,824.0,,36,1,11,6,512.033,512.033,0,3,0,0,0,0,0,0,0,Larger than 6
491,Calcium dobesilate,418.4,212.0,228.0,,25,4,10,0,417.934,417.934,0,3,0,0,0,0,0,0,0,Larger than 6
492,Amoxicillin,365.4,158.0,590.0,-2.0,25,4,7,4,365.105,365.105,0,1,0,4,4,0,0,0,0,Between -11 and 5
493,Tramadol,263.37,32.7,282.0,2.6,19,1,3,4,263.189,263.189,0,1,0,2,2,0,0,0,0,Between -11 and 5
494,Almitrine,477.6,69.2,602.0,5.6,35,2,9,10,477.245,477.245,0,1,0,0,0,0,0,0,0,Larger than 6
495,Pirfenidone,185.22,20.3,285.0,1.9,14,0,1,1,185.084,185.084,0,1,0,0,0,0,0,0,0,Between -11 and 5
496,Piperacillin,517.6,182.0,982.0,0.5,36,3,8,6,517.163,517.163,0,1,0,4,4,0,0,0,0,Between -11 and 5
497,Captopril,217.29,58.6,244.0,0.3,14,2,4,3,217.077,217.077,0,1,0,2,2,0,0,0,0,Between -11 and 5
498,Nicorandil,211.17,97.0,228.0,0.8,15,1,5,4,211.059,211.059,0,1,0,0,0,0,0,0,0,Between -11 and 5
499,Brequinar,375.4,50.2,551.0,5.6,28,1,5,3,375.107,375.107,0,1,0,0,0,0,0,0,0,Larger than 6
500,Amantadine hydrochloride,187.71,26.0,144.0,,12,2,1,0,187.113,187.113,0,2,0,0,0,0,0,0,0,Larger than 6
501,Chloroquine Phosphate,515.9,184.0,359.0,,32,7,11,8,515.135,515.135,0,3,0,1,0,1,0,0,0,Larger than 6
502,Epigallocatechin Gallate,458.4,197.0,667.0,1.2,33,8,11,4,458.085,458.085,0,1,0,2,2,0,0,0,0,Between -11 and 5
503,Artemisinin,282.33,54.0,452.0,2.8,20,0,5,0,282.147,282.147,0,1,0,7,7,0,0,0,0,Between -11 and 5
504,"1-((2S,3R,4S,5S)-3-Fluoro-4-hydroxy-5-(hydroxymethyl)tetrahydrofuran-2-yl)-5-methylpyrimidine-2,4(1H,3H)-dione",260.22,99.1,413.0,-0.9,18,3,6,2,260.081,260.081,0,1,0,4,4,0,0,0,0,Between -11 and 5
505,Bis (2-Carboxyethylgermanium)sesquioxide,339.4,118.0,259.0,,15,2,7,8,339.886,341.885,0,1,0,0,0,0,0,0,0,Larger than 6
506,Clarithromycin,748.0,183.0,1190.0,3.2,52,4,14,8,747.477,747.477,0,1,0,18,18,0,0,0,0,Between -11 and 5
507,"2,6-Difluorophenol",130.09,20.2,87.1,2.0,9,1,3,0,130.023,130.023,0,1,0,0,0,0,0,0,0,Between -11 and 5
508,Moxifloxacin Hydrochloride,437.9,82.1,727.0,,30,3,8,4,437.152,437.152,0,2,0,2,2,0,0,0,0,Larger than 6
509,Levofloxacin,361.4,73.3,634.0,-0.4,26,1,8,2,361.144,361.144,0,1,0,1,1,0,0,0,0,Between -11 and 5
510,Olmesartan,446.5,130.0,656.0,3.2,33,3,7,8,446.207,446.207,0,1,0,0,0,0,0,0,0,Between -11 and 5
511,Tetrasul sulfoxide,340.0,36.3,307.0,5.2,18,0,2,2,339.886,337.889,0,1,0,1,0,1,0,0,0,Larger than 6
512,Ramelteon,259.339,38.3,331.0,2.7,19,1,2,4,259.157,259.157,0,1,0,1,1,0,0,0,0,Between -11 and 5
513,Darunavir,547.7,149.0,853.0,2.9,38,3,9,12,547.235,547.235,0,1,0,5,5,0,0,0,0,Between -11 and 5
514,Galactomannan,504.4,269.0,641.0,-6.3,34,11,16,7,504.169,504.169,0,1,0,15,15,0,0,0,0,Between -11 and 5
515,Mometasone furoate,521.4,93.8,1020.0,3.9,35,1,6,5,520.142,520.142,0,1,0,8,8,0,0,0,0,Between -11 and 5
516,Rosuvastatin,481.5,149.0,767.0,1.6,33,3,10,10,481.168,481.168,0,1,0,2,2,0,1,1,0,Between -11 and 5
517,Tenofovir,287.21,136.0,354.0,-1.6,19,3,8,5,287.078,287.078,0,1,0,1,1,0,0,0,0,Between -11 and 5
518,3-Cyclopentyl-1-(piperazin-1-yl)propan-1-one,210.32,32.299,206.0,1.6,15,1,2,3,210.173,210.173,0,1,0,0,0,0,0,0,0,Between -11 and 5
519,Azithromycin Dihydrate,785.0,182.0,1150.0,,54,7,16,7,784.53,784.53,0,3,0,18,18,0,0,0,0,Larger than 6
520,"D-Alanine, N-methylglycyl-L-arginyl-L-valyl-L-tyrosyl-L-isoleucyl-L-histidyl-L-prolyl-",926.1,358.0,1690.0,-2.3,66,12,13,26,925.513,925.513,0,1,0,8,8,0,0,0,0,Between -11 and 5
521,Arformoterol,344.4,90.8,388.0,1.8,25,4,5,8,344.174,344.174,0,1,0,2,2,0,0,0,0,Between -11 and 5
522,3-Hydroxybutyrate,103.1,60.4,63.8,0.1,7,1,3,1,103.04,103.04,-1,1,0,1,0,1,0,0,0,Between -11 and 5
523,Sitagliptin,407.31,77.0,566.0,0.7,28,1,10,4,407.118,407.118,0,1,0,1,1,0,0,0,0,Between -11 and 5
524,Best,572.3,101.0,412.0,,18,0,4,0,573.618,575.617,0,1,0,0,0,0,0,0,0,Larger than 6
525,Eprosartan,424.5,121.0,618.0,4.5,30,2,6,10,424.146,424.146,0,1,0,0,0,0,1,1,0,Between -11 and 5
526,Isotretinoin,300.4,37.3,567.0,6.3,22,1,2,5,300.209,300.209,0,1,0,0,0,0,4,4,0,Larger than 6
527,Epoprostenol,352.5,87.0,485.0,2.9,25,3,5,10,352.225,352.225,0,1,0,5,5,0,2,2,0,Between -11 and 5
528,Camostat mesylate,494.5,200.0,695.0,,34,3,9,9,494.147,494.147,0,2,0,0,0,0,0,0,0,Larger than 6
529,Fluticasone,444.5,99.9,861.0,3.2,30,2,8,3,444.158,444.158,0,1,0,9,9,0,0,0,0,Between -11 and 5
530,Tenofovir disoproxil,519.4,185.0,698.0,1.6,35,1,14,17,519.173,519.173,0,1,0,1,1,0,0,0,0,Between -11 and 5
531,Refanalin,176.24,56.9,170.0,2.2,12,1,2,2,176.041,176.041,0,1,0,0,0,0,1,1,0,Between -11 and 5
532,Sulodexide,295.29,115.0,363.0,0.2,21,3,8,4,295.128,295.128,0,1,0,4,4,0,0,0,0,Between -11 and 5
533,Metampicillin,361.4,124.0,603.0,3.0,25,2,6,5,361.11,361.11,0,1,0,4,4,0,0,0,0,Between -11 and 5
534,Ciclesonide,540.7,99.1,1100.0,5.3,39,1,7,6,540.309,540.309,0,1,0,9,9,0,0,0,0,Larger than 6
535,Dutasteride,528.5,58.2,964.0,5.4,37,2,8,2,528.221,528.221,0,1,0,7,7,0,0,0,0,Larger than 6
536,Prasugrel,373.4,74.8,555.0,3.6,26,0,6,6,373.115,373.115,0,1,0,1,0,1,0,0,0,Between -11 and 5
537,Almitrine mesylate,669.8,195.0,694.0,,45,4,15,10,669.221,669.221,0,3,0,0,0,0,0,0,0,Larger than 6
538,Ile-Ser,218.25,113.0,232.0,-3.3,15,4,5,6,218.127,218.127,0,1,0,3,3,0,0,0,0,Between -11 and 5
539,butyl (3-(4-((1H-imidazol-1-yl)methyl)phenyl)-5-isobutylthiophen-2-yl)sulfonylcarbamate,475.6,127.0,690.0,5.3,32,1,6,11,475.16,475.16,0,1,0,0,0,0,0,0,0,Larger than 6
540,Vidofludimus,355.4,75.6,576.0,3.4,26,2,5,5,355.122,355.122,0,1,0,0,0,0,0,0,0,Between -11 and 5
541,Reparixin,283.39,71.6,389.0,2.9,19,1,3,5,283.124,283.124,0,1,0,1,1,0,0,0,0,Between -11 and 5
542,Ticagrelor,522.6,164.0,736.0,2.0,36,4,12,10,522.186,522.186,0,1,0,6,6,0,0,0,0,Between -11 and 5
543,Anhydrous Ceftriaxone Sodium,577.6,288.0,1110.0,,37,4,13,8,577.036,577.036,0,2,0,2,2,0,1,1,0,Larger than 6
544,Tradipitant,587.9,73.6,865.0,6.2,41,0,11,6,587.095,587.095,0,1,0,0,0,0,0,0,0,Larger than 6
545,CID 9939931,701.6,316.0,819.0,,47,12,9,18,700.261,700.261,0,2,0,4,4,0,2,2,0,Larger than 6
546,1-Palmitoyl-2-linoleoyl-3-acetyl-rac-glycerol,635.0,78.9,744.0,14.1,45,0,6,36,634.517,634.517,0,1,0,1,0,1,2,2,0,Larger than 6
547,Zinforo,744.7,368.0,1240.0,,47,5,19,10,744.031,744.031,0,2,0,2,2,0,1,1,0,Larger than 6
548,Linagliptin,472.5,114.0,885.0,1.9,35,1,7,4,472.234,472.234,0,1,0,1,1,0,0,0,0,Between -11 and 5
549,"2-(2-Chloro-4-iodophenylamino)-3,4-difluorobenzoic acid",409.55,49.3,363.0,5.0,20,2,5,3,408.918,408.918,0,1,0,0,0,0,0,0,0,Between -11 and 5
550,Edoxaban,548.1,165.0,880.0,1.4,37,3,8,5,547.177,547.177,0,1,0,3,3,0,0,0,0,Between -11 and 5
551,Transcrocetinate sodium,372.4,80.3,597.0,,26,0,4,6,372.131,372.131,0,3,0,0,0,0,7,7,0,Larger than 6
552,Galidesivir,265.27,140.0,334.0,-2.1,19,6,7,2,265.117,265.117,0,1,0,4,4,0,0,0,0,Between -11 and 5
553,"4,5-Dihydro-3-phenyl-5-isoxazoleacetic acid",205.21,58.9,269.0,1.3,15,1,4,3,205.074,205.074,0,1,0,1,0,1,0,0,0,Between -11 and 5
554,5-Cholesten-3beta-25-diol-3-sulfate,482.7,92.2,858.0,6.4,33,2,5,7,482.307,482.307,0,1,0,8,8,0,0,0,0,Larger than 6
555,"disodium;(6R,7R)-7-[[(2Z)-2-(2-amino-1,3-thiazol-4-yl)-2-methoxyiminoacetyl]amino]-3-[(2-methyl-6-oxido-5-oxo-1,2,4-triazin-3-yl)sulfanylmethyl]-8-oxo-5-thia-1-azabicyclo[4.2.0]oct-2-ene-2-carboxylate;hydrate",616.6,298.0,1120.0,,39,3,15,7,616.021,616.021,0,4,0,2,2,0,1,1,0,Larger than 6
556,Losartan potassium,461.0,77.7,526.0,,31,1,6,8,460.118,460.118,0,2,0,0,0,0,0,0,0,Larger than 6
557,Lopinavir and ritonavir,1349.7,322.0,1980.0,,96,8,14,33,1348.68,1348.68,0,2,0,8,8,0,0,0,0,Larger than 6
558,"N-[5-[[2-(2,6-dimethylphenoxy)acetyl]amino]-4-hydroxy-1,6-diphenylhexan-2-yl]-3-methyl-2-(2-oxo-1,3-diazinan-1-yl)butanamide;1,3-thiazol-5-ylmethyl N-[(2S,3S,5S)-3-hydroxy-5-[[(2S)-3-methyl-2-[[methyl-[(2-propan-2-yl-1,3-thiazol-4-yl)methyl]carbamoyl]amino]butanoyl]amino]-1,6-diphenylhexan-2-yl]carbamate",1349.7,322.0,1980.0,,96,8,14,33,1348.68,1348.68,0,2,0,8,4,4,0,0,0,Larger than 6
559,"S-[2-[3-[[(2R)-4-[[[(2R,3S,4R,5R)-5-(6-aminopurin-9-yl)-4-hydroxy-3-phosphonooxyoxolan-2-yl]methoxy-hydroxyphosphoryl]oxy-hydroxyphosphoryl]oxy-2-hydroxy-3,3-dimethylbutanoyl]amino]propanoylamino]ethyl] (2S)-2-[4-(2-methylpropyl)phenyl]propanethioate",955.8,389.0,1660.0,-1.9,62,9,22,24,955.235,955.235,0,1,0,6,6,0,0,0,0,Between -11 and 5
560,Bivalirudin,2180.3,902.0,4950.0,-7.1,155,28,35,67,2179.99,2178.99,0,1,0,16,16,0,0,0,0,Between -11 and 5
561,Thymalfasin,3108.3,1460.0,7190.0,-24.0,217,49,59,111,3107.51,3106.5,0,1,0,32,32,0,0,0,0,Smaller than -10
562,CID 16219160,397.6,74.6,707.0,,27,2,4,3,397.178,397.178,0,2,0,6,6,0,0,0,0,Larger than 6
563,Sodium valproate,166.19,40.1,98.3,,11,0,2,5,166.097,166.097,0,2,0,0,0,0,0,0,0,Larger than 6
564,"N,N-Diethyl-(2-(4-(2-(18F)fluoroethoxy)phenyl)-5,7-dimethylpyrazolo(1,5-A)pyrimidine-3-YL)acetamide",397.5,59.7,525.0,3.2,29,0,5,8,397.214,397.214,0,1,1,0,0,0,0,0,0,Between -11 and 5
565,Amoxicillin sodium,387.4,161.0,596.0,,26,3,7,4,387.086,387.086,0,2,0,4,4,0,0,0,0,Larger than 6
566,Potassium canrenoate,396.6,77.4,713.0,,27,1,4,3,396.17,396.17,0,2,0,6,6,0,0,0,0,Larger than 6
567,Hydrocortisone sodium succinate,484.5,141.0,915.0,,34,2,8,7,484.207,484.207,0,2,0,7,7,0,0,0,0,Larger than 6
568,Piperacillin-tazobactam,839.8,315.0,1560.0,,57,3,15,9,839.198,839.198,0,3,0,7,7,0,0,0,0,Larger than 6
569,Antroquinonol,390.6,55.8,648.0,5.8,28,1,4,10,390.277,390.277,0,1,0,3,3,0,2,2,0,Larger than 6
570,Brilacidin,936.9,314.0,1560.0,0.3,66,10,18,20,936.394,936.394,0,1,0,2,2,0,0,0,0,Between -11 and 5
571,Ruxolitinib phosphate,404.4,161.0,503.0,,28,4,8,4,404.136,404.136,0,2,0,1,1,0,0,0,0,Larger than 6
572,Baricitinib,371.4,129.0,678.0,-0.5,26,1,7,5,371.116,371.116,0,1,0,0,0,0,0,0,0,Between -11 and 5
573,"(2R,3R,4S,5R)-2-(4-aminopyrrolo[2,1-f][1,2,4]triazin-7-yl)-3,4-dihydroxy-5-(hydroxymethyl)tetrahydrofuran-2-carbonitrile",291.26,150.0,456.0,-1.4,21,4,8,2,291.097,291.097,0,1,0,4,4,0,0,0,0,Between -11 and 5
574,Metformin glycinate,204.23,155.0,175.0,,14,5,4,3,204.133,204.133,0,2,0,0,0,0,0,0,0,Larger than 6
575,1-Palmityl-2-(4-carboxybutyl)-SN-glycero-3-phosphocholine,581.8,114.0,606.0,6.7,39,1,8,30,581.406,581.406,0,1,0,1,1,0,0,0,0,Larger than 6
576,3-phenyl-4-propyl-1-(pyridin-2-yl)-1H-pyrazol-5-ol,279.34,45.2,412.0,3.7,21,1,3,4,279.137,279.137,0,1,0,0,0,0,0,0,0,Between -11 and 5
577,Ozanimod,404.5,104.0,609.0,3.1,30,2,7,7,404.185,404.185,0,1,0,1,1,0,0,0,0,Between -11 and 5
578,Sabizabulin,377.4,89.2,534.0,3.4,28,2,5,6,377.138,377.138,0,1,0,0,0,0,0,0,0,Between -11 and 5
579,Zavegepant,638.8,117.0,1160.0,3.1,47,3,6,6,638.369,638.369,0,1,0,1,1,0,0,0,0,Between -11 and 5
580,CID 53477736,749.0,180.0,1150.0,4.0,52,5,14,7,748.509,748.509,0,1,0,18,18,0,0,0,0,Between -11 and 5
581,Danoprevir (RG7227),731.8,189.0,1530.0,3.3,51,3,10,8,731.3,731.3,0,1,0,5,5,0,1,1,0,Between -11 and 5
582,Solu-Medrol,497.5,138.0,981.0,,35,3,8,7,497.215,497.215,1,2,0,8,8,0,0,0,0,Larger than 6
583,Vericiguat,426.4,147.0,622.0,1.5,31,3,10,5,426.136,426.136,0,1,0,0,0,0,0,0,0,Between -11 and 5
584,Aldose reductase-IN-1,421.4,137.0,706.0,2.3,29,1,11,4,421.046,421.046,0,1,0,0,0,0,0,0,0,Between -11 and 5
585,Ixazomib citrate,517.1,168.0,797.0,,34,4,9,11,516.087,516.087,0,1,0,1,1,0,0,0,0,Larger than 6
586,Vidofludimus calcium anhydrous,748.8,157.0,571.0,,53,2,10,8,748.191,748.191,0,3,0,0,0,0,0,0,0,Larger than 6
587,Upadacitinib,380.4,78.3,561.0,2.7,27,2,6,3,380.157,380.157,0,1,0,2,2,0,0,0,0,Between -11 and 5
588,Asapiprant,501.6,131.0,789.0,3.1,35,1,10,9,501.157,501.157,0,1,0,0,0,0,0,0,0,Between -11 and 5
589,"1,1'-hexamethylene bis[5-(p-chlorophenyl) biguanide] di-D-gluconate",735.7,317.0,819.0,,48,13,10,18,734.249,734.249,0,3,0,4,4,0,2,2,0,Larger than 6
590,Proxalutamide,517.5,118.0,894.0,4.3,36,0,10,6,517.12,517.12,0,1,0,0,0,0,0,0,0,Between -11 and 5
591,Rocefin,652.6,300.0,1120.0,,41,5,17,7,652.042,652.042,0,6,0,2,2,0,1,1,0,Larger than 6
592,Galidesivir hydrochloride,301.73,140.0,334.0,,20,7,7,2,301.094,301.094,0,2,0,4,4,0,0,0,0,Larger than 6
593,Entresto,1916.0,396.0,1140.0,,135,7,29,40,1915.81,1914.81,0,15,0,6,6,0,0,0,0,Larger than 6
594,"(2R,6S,12Z,13aS,14aR,16aS)-6-[(tert-Butoxycarbonyl)amino]-14a-[(cyclopropylsulfonyl)carbamoyl]-5,16-dioxo-1,2,3,5,6,7,8,9,10,11,13a,14,14a,15,16,16a-hexadecahydrocyclopropa[e]pyrrolo[1,2-a][1,4]diazacyclopentadecin-2-yl 4-fluoro-1,3-dihydro-2H-isoindole-2-carboxylate",731.8,189.0,1530.0,3.3,51,3,10,8,731.3,731.3,0,1,0,5,5,0,1,1,0,Between -11 and 5
595,Harvoni,1418.4,327.0,2730.0,,101,7,21,23,1417.58,1417.58,0,2,0,12,12,0,0,0,0,Larger than 6
596,Abivertinib,487.5,98.4,752.0,4.2,36,3,8,7,487.213,487.213,0,1,0,0,0,0,0,0,0,Between -11 and 5
597,Legalon SIL,1453.1,495.0,1200.0,,102,6,32,24,1452.23,1452.23,0,6,0,8,8,0,0,0,0,Larger than 6
598,Maltofer,449.16,200.0,367.0,,27,11,14,8,449.059,449.059,0,5,0,9,9,0,0,0,0,Larger than 6
599,Zunsemetinib,513.9,101.0,888.0,2.5,36,1,9,6,513.138,513.138,0,1,0,0,0,0,0,0,0,Between -11 and 5
600,Treamid,318.37,116.0,345.0,-0.7,23,4,4,10,318.18,318.18,0,1,0,0,0,0,0,0,0,Between -11 and 5
601,CID 87071853,1003.2,299.0,767.0,,67,6,20,20,1002.3,1002.3,0,3,0,4,4,0,2,2,0,Larger than 6
602,"disodium;[2-[(8S,9R,10S,11S,13S,14S,16R,17R)-9-fluoro-11,17-dihydroxy-10,13,16-trimethyl-3-oxo-6,7,8,11,12,14,15,16-octahydrocyclopenta[a]phenanthren-17-yl]-2-oxoethyl] phosphate;hydrate",534.4,148.0,962.0,,35,3,10,3,534.141,534.141,0,4,0,8,8,0,0,0,0,Larger than 6
603,Heparin sodium,1157.9,652.0,2410.0,,71,15,38,21,1157.0,1157.0,1,2,0,20,0,20,0,0,0,Larger than 6
604,CID 101731853,2088.6,933.0,4400.0,3.1,140,40,48,73,2086.96,2086.96,0,1,0,18,16,2,0,0,0,Between -11 and 5
605,"N-(5-Oxidanyl-1,3-Benzothiazol-2-Yl)ethanamide",208.24,90.5,237.0,1.5,14,2,4,1,208.031,208.031,0,1,0,0,0,0,0,0,0,Between -11 and 5
606,Danicopan,580.4,123.0,891.0,3.3,38,1,8,6,579.103,579.103,0,1,0,2,2,0,0,0,0,Between -11 and 5
607,"Disodium;2-[[2-[[4-(2,2-dimethylpropanoyloxy)phenyl]sulfonylamino]benzoyl]amino]acetate",479.4,150.0,726.0,,32,2,8,8,479.086,479.086,1,3,0,0,0,0,0,0,0,Larger than 6
608,"propan-2-yl (2S)-2-[[[(3R,4R,5R)-5-[2-amino-6-(methylamino)purin-9-yl]-4-fluoro-3-hydroxy-4-methyloxolan-2-yl]methoxy-phenoxyphosphoryl]amino]propanoate",581.5,185.0,919.0,1.7,40,4,14,12,581.216,581.216,0,1,0,6,5,1,0,0,0,Between -11 and 5
609,"[(1S,4R,6S,7E,18R)-4-(cyclopropylsulfonylcarbamoyl)-14-[(2-methylpropan-2-yl)oxycarbonylamino]-2,15-dioxo-3,16-diazatricyclo[14.3.0.04,6]nonadec-7-en-18-yl] 4-fluoro-1,3-dihydroisoindole-2-carboxylate",731.8,189.0,1530.0,3.3,51,3,10,8,731.3,731.3,0,1,0,5,4,1,1,1,0,Between -11 and 5
610,Eclitasertib,378.4,113.0,570.0,1.7,28,2,6,4,378.144,378.144,0,1,0,1,1,0,0,0,0,Between -11 and 5
611,Dazcapistat,395.4,115.0,611.0,2.9,29,2,6,7,395.128,395.128,0,1,0,1,0,1,0,0,0,Between -11 and 5
612,Bexotegrast,492.6,113.0,655.0,1.8,36,3,9,14,492.285,492.285,0,1,0,1,1,0,0,0,0,Between -11 and 5
613,Estetrol monohydrate,322.4,81.9,441.0,,23,5,5,0,322.178,322.178,0,2,0,7,7,0,0,0,0,Larger than 6
614,Sildenafil,474.6,118.0,838.0,1.5,33,1,8,7,474.205,474.205,0,1,0,0,0,0,0,0,0,Between -11 and 5
615,Azilsartan,456.4,115.0,783.0,4.4,34,2,7,7,456.143,456.143,0,1,0,0,0,0,0,0,0,Between -11 and 5
616,Echinochrome A,266.2,135.0,455.0,2.0,19,5,7,1,266.043,266.043,0,1,0,0,0,0,0,0,0,Between -11 and 5
617,F-Arag F-18,284.23,135.0,449.0,-0.9,20,4,7,2,284.09,284.09,0,1,1,4,4,0,0,0,0,Between -11 and 5
618,Apabetalone,370.4,89.4,543.0,2.3,27,2,6,6,370.153,370.153,0,1,0,0,0,0,0,0,0,Between -11 and 5
619,"4-acetamidobenzoic acid;9-[(2R,3R,4S,5R)-3,4-dihydroxy-5-(hydroxymethyl)oxolan-2-yl]-1H-purin-6-one;(2R)-1-(dimethylamino)propan-2-ol",1115.2,399.0,658.0,,79,13,22,14,1114.55,1114.55,0,7,0,7,7,0,0,0,0,Larger than 6
620,"[(1R,3R)-4-[(15Z,17E)-16-formyl-18-(4-hydroxy-2,2,6,6-tetramethylcyclohexyl)-3,7,12-trimethyl-14-oxooctadeca-2,4,6,8,10,12,15,17-octaenylidene]-3-hydroxy-3,5,5-trimethylcyclohexyl] acetate",672.9,101.0,1480.0,9.3,49,2,6,13,672.439,672.439,0,1,0,2,2,0,9,2,7,Larger than 6
621,Abivertinib maleate,639.6,175.0,871.0,,46,7,14,9,639.245,639.245,0,4,0,0,0,0,1,1,0,Larger than 6
622,"7-[[2-(2-Amino-1,3-thiazol-4-yl)-2-(2,2-dimethylpropanoyloxymethoxyimino)acetyl]amino]-3-ethenyl-8-oxo-5-thia-1-azabicyclo[4.2.0]oct-2-ene-2-carboxylic acid",509.6,227.0,961.0,1.1,34,3,12,10,509.104,509.104,0,1,0,2,0,2,1,0,1,Between -11 and 5
623,Unii-T5UX5skk2S,452.5,120.0,723.0,3.5,33,4,4,9,452.242,452.242,0,1,0,3,3,0,0,0,0,Between -11 and 5
624,"azane;(2R,3S,4S,5R,6R)-2-(hydroxymethyl)-6-[(2R,3S,4R,5R,6S)-4,5,6-trihydroxy-2-(hydroxymethyl)oxan-3-yl]oxyoxane-3,4,5-triol",359.33,191.0,382.0,,24,9,12,4,359.143,359.143,0,2,0,10,10,0,0,0,0,Larger than 6
625,Nezulcitinib,527.7,104.0,866.0,3.6,39,3,6,6,527.301,527.301,0,1,0,1,1,0,0,0,0,Between -11 and 5
626,P9Zqs28F8C,403.5,86.9,712.0,3.6,28,2,4,4,403.193,403.193,0,1,0,1,1,0,1,1,0,Between -11 and 5
627,Lufotrelvir,552.5,196.0,927.0,0.5,38,6,9,13,552.199,552.199,0,1,0,3,3,0,0,0,0,Between -11 and 5
628,Nirmatrelvir,499.5,131.0,964.0,2.2,35,3,8,7,499.241,499.241,0,1,0,6,6,0,0,0,0,Between -11 and 5
629,"hexasodium;4-[[(2S,4R)-5-ethoxy-4-methyl-5-oxo-1-(4-phenylphenyl)pentan-2-yl]amino]-4-oxobutanoate;hydride;(2S)-3-methyl-2-[pentanoyl-[[4-[2-(1,2,3-triaza-4-azanidacyclopenta-2,5-dien-5-yl)phenyl]phenyl]methyl]amino]butanoate;pentahydrate",1922.0,396.0,1140.0,,135,7,35,40,1921.86,1920.85,-6,21,0,6,6,0,0,0,0,Larger than 6
630,Nangibotide,1342.5,634.0,2630.0,-8.3,92,20,24,45,1341.53,1341.53,0,1,0,10,10,0,0,0,0,Between -11 and 5
Back to top ]]> Python Shiny Pandas Plotly PubChem Cheminformatics https://jhylin.github.io/Data_in_life_blog/posts/13_Shiny_app_python/ShinyAppPy_PC_Cov19_app_embed.html Sun, 07 May 2023 12:00:00 GMT Shinylive app in Python Jennifer HY Lin https://jhylin.github.io/Data_in_life_blog/posts/13_Shiny_app_python/ShinyAppPy_PC_Cov19_data_prep.html
Brief introduction

Since I’ve had a lot of fun building a Shiny app in R last time, I was on track to build another Shiny app again but using Python instead. So here in this post, I’ll talk about the data wrangling process to prepare the final dataset needed to build a Shinylive app in Python. The actual Shinylive app deployment and access will be shown in a separate post after this one.


Source of data

The dataset used for this Shiny app in Python was from PubChem (link here). There were a total of 631 compounds at the time when I downloaded them as .csv file, along with their relevant compound data. I only picked this dataset randomly, as the focus would be more on app building, but it was nice to see an interactive web app being built and used for a domain such as pharmaceutical research.


Import Polars

Polars dataframe library was used again this time.

import polars as pl


Reading .csv file
pc = pl.read_csv("pubchem.csv")
pc.head()
shape: (5, 38)
cid cmpdname cmpdsynonym mw mf polararea complexity xlogp heavycnt hbonddonor hbondacc rotbonds inchi isosmiles canonicalsmiles inchikey iupacname exactmass monoisotopicmass charge covalentunitcnt isotopeatomcnt totalatomstereocnt definedatomstereocnt undefinedatomstereocnt totalbondstereocnt definedbondstereocnt undefinedbondstereocnt pclidcnt gpidcnt meshheadings annothits annothitcnt aids cidcdate sidsrcname depcatg annotation
i64 str str f64 str f64 f64 str i64 i64 i64 i64 str str str str str f64 f64 i64 i64 i64 i64 i64 i64 i64 i64 i64 i64 i64 str str i64 str i64 str str str
5280453 "Calcitriol" "calcitriol|322... 416.6 "C27H44O3" 60.7 688.0 "5.100" 30 3 3 6 "InChI=1S/C27H4... "C[C@H](CCCC(C)... "CC(CCCC(C)(C)O... "GMRQFYUYWCNGIN... "(1R,3S,5Z)-5-[... 416.329 416.329 0 1 0 6 6 0 2 2 0 22311 46029 "Calcitriol" "Biological Tes... 12 "485|631|731|78... 20040916 "A2B Chem|AA BL... "Chemical Vendo... "COVID-19, COVI...
9962735 "Ubiquinol" "ubiquinol|992-... 865.4 "C59H92O4" 58.9 1600.0 "20.200" 63 2 4 31 "InChI=1S/C59H9... "CC1=C(C(=C(C(=... "CC1=C(C(=C(C(=... "QNTNKSLOFHEFPK... "2-[(2E,6E,10E,... 864.7 864.7 0 1 0 0 0 0 9 9 0 2732 21358 "NULL" "Chemical and P... 7 "NULL" 20061025 "001Chemical|A2... "Chemical Vendo... "COVID-19, COVI...
5961 "Glutamine" "L-glutamine|gl... 146.14 "C5H10N2O3" 106.0 146.0 "-3.100" 10 3 4 4 "InChI=1S/C5H10... "C(CC(=O)N)[C@@... "C(CC(=O)N)C(C(... "ZDXPYRJPNDTMRX... "(2S)-2,5-diami... 146.069 146.069 0 1 0 1 1 0 0 0 0 88218 399 "Glutamine" "Biological Tes... 12 "422|429|436|54... 20040916 "001Chemical|3B... "Chemical Vendo... "COVID-19, COVI...
2244 "Aspirin" "aspirin|ACETYL... 180.16 "C9H8O4" 63.6 212.0 "1.200" 13 1 4 3 "InChI=1S/C9H8O... "CC(=O)OC1=CC=C... "CC(=O)OC1=CC=C... "BSYNRYMUTXBXSQ... "2-acetyloxyben... 180.042 180.042 0 1 0 0 0 0 0 0 0 127012 364455 "Aspirin" "Biological Tes... 12 "1|3|9|15|19|21... 20040916 "001Chemical|3B... "Chemical Vendo... "COVID-19, COVI...
457 "1-Methylnicoti... "1-methylnicoti... 137.16 "C7H9N2O+" 47.0 136.0 "-0.100" 10 1 1 1 "InChI=1S/C7H8N... "C[N+]1=CC=CC(=... "C[N+]1=CC=CC(=... "LDHMAVIPBRSVRG... "1-methylpyridi... 137.071 137.071 1 1 0 0 0 0 0 0 0 310 674 "NULL" "Biological Tes... 8 "61001|61002|14... 20040916 "001Chemical|3B... "Chemical Vendo... "COVID-19, COVI...


Quick look at the data

I decided to comment out the code below to keep the post at a reasonable length for reading purpose, but they were very handy for a quick glimpse of the data content.

# Quick overview of the variables in each column in the dataset
# Uncomment line below if needed to run
#print(pc.glimpse())

# Quick look at all column names
# Uncomment line below if needed to run
#pc.columns


Check for nulls in dataset
pc.null_count()
shape: (1, 38)
cid cmpdname cmpdsynonym mw mf polararea complexity xlogp heavycnt hbonddonor hbondacc rotbonds inchi isosmiles canonicalsmiles inchikey iupacname exactmass monoisotopicmass charge covalentunitcnt isotopeatomcnt totalatomstereocnt definedatomstereocnt undefinedatomstereocnt totalbondstereocnt definedbondstereocnt undefinedbondstereocnt pclidcnt gpidcnt meshheadings annothits annothitcnt aids cidcdate sidsrcname depcatg annotation
u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32 u32
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0


Change column names as needed
# Change column names
pc_cov = pc.rename(
    {
        "cmpdname": "Compound name",
        "cmpdsynonym": "Synonyms",
        "mw": "Molecular weight",
        "mf": "Molecular formula",
        "polararea": "Polar surface area",
        "complexity": "Complexity",
        "xlogp": "Partition coefficients",
        "heavycnt": "Heavy atom count",
        "hbonddonor": "Hydrogen bond donor count",
        "hbondacc": "Hydrogen bond acceptor count",
        "rotbonds": "Rotatable bond count",
        "exactmass": "Exact mass",
        "monoisotopicmass": "Monoisotopic mass",
        "charge": "Formal charge",
        "covalentunitcnt": "Covalently-bonded unit count",
        "isotopeatomcnt": "Isotope atom count",
        "totalatomstereocnt": "Total atom stereocenter count",
        "definedatomstereocnt": "Defined atom stereocenter count",
        "undefinedatomstereocnt": "Undefined atoms stereocenter count",
        "totalbondstereocnt": "Total bond stereocenter count",
        "definedbondstereocnt": "Defined bond stereocenter count",
        "undefinedbondstereocnt": "Undefined bond stereocenter count",
        "meshheadings": "MeSH headings"
    }
)

pc_cov.head()
shape: (5, 38)
cid Compound name Synonyms Molecular weight Molecular formula Polar surface area Complexity Partition coefficients Heavy atom count Hydrogen bond donor count Hydrogen bond acceptor count Rotatable bond count inchi isosmiles canonicalsmiles inchikey iupacname Exact mass Monoisotopic mass Formal charge Covalently-bonded unit count Isotope atom count Total atom stereocenter count Defined atom stereocenter count Undefined atoms stereocenter count Total bond stereocenter count Defined bond stereocenter count Undefined bond stereocenter count pclidcnt gpidcnt MeSH headings annothits annothitcnt aids cidcdate sidsrcname depcatg annotation
i64 str str f64 str f64 f64 str i64 i64 i64 i64 str str str str str f64 f64 i64 i64 i64 i64 i64 i64 i64 i64 i64 i64 i64 str str i64 str i64 str str str
5280453 "Calcitriol" "calcitriol|322... 416.6 "C27H44O3" 60.7 688.0 "5.100" 30 3 3 6 "InChI=1S/C27H4... "C[C@H](CCCC(C)... "CC(CCCC(C)(C)O... "GMRQFYUYWCNGIN... "(1R,3S,5Z)-5-[... 416.329 416.329 0 1 0 6 6 0 2 2 0 22311 46029 "Calcitriol" "Biological Tes... 12 "485|631|731|78... 20040916 "A2B Chem|AA BL... "Chemical Vendo... "COVID-19, COVI...
9962735 "Ubiquinol" "ubiquinol|992-... 865.4 "C59H92O4" 58.9 1600.0 "20.200" 63 2 4 31 "InChI=1S/C59H9... "CC1=C(C(=C(C(=... "CC1=C(C(=C(C(=... "QNTNKSLOFHEFPK... "2-[(2E,6E,10E,... 864.7 864.7 0 1 0 0 0 0 9 9 0 2732 21358 "NULL" "Chemical and P... 7 "NULL" 20061025 "001Chemical|A2... "Chemical Vendo... "COVID-19, COVI...
5961 "Glutamine" "L-glutamine|gl... 146.14 "C5H10N2O3" 106.0 146.0 "-3.100" 10 3 4 4 "InChI=1S/C5H10... "C(CC(=O)N)[C@@... "C(CC(=O)N)C(C(... "ZDXPYRJPNDTMRX... "(2S)-2,5-diami... 146.069 146.069 0 1 0 1 1 0 0 0 0 88218 399 "Glutamine" "Biological Tes... 12 "422|429|436|54... 20040916 "001Chemical|3B... "Chemical Vendo... "COVID-19, COVI...
2244 "Aspirin" "aspirin|ACETYL... 180.16 "C9H8O4" 63.6 212.0 "1.200" 13 1 4 3 "InChI=1S/C9H8O... "CC(=O)OC1=CC=C... "CC(=O)OC1=CC=C... "BSYNRYMUTXBXSQ... "2-acetyloxyben... 180.042 180.042 0 1 0 0 0 0 0 0 0 127012 364455 "Aspirin" "Biological Tes... 12 "1|3|9|15|19|21... 20040916 "001Chemical|3B... "Chemical Vendo... "COVID-19, COVI...
457 "1-Methylnicoti... "1-methylnicoti... 137.16 "C7H9N2O+" 47.0 136.0 "-0.100" 10 1 1 1 "InChI=1S/C7H8N... "C[N+]1=CC=CC(=... "C[N+]1=CC=CC(=... "LDHMAVIPBRSVRG... "1-methylpyridi... 137.071 137.071 1 1 0 0 0 0 0 0 0 310 674 "NULL" "Biological Tes... 8 "61001|61002|14... 20040916 "001Chemical|3B... "Chemical Vendo... "COVID-19, COVI...


Definitions of molecular properties in this PubChem dataset

The definitions for some of the column names were shown below, which were mainly derived and adapted from PubChem:

Note: please refer to PubChem documentations for full definitions

  • Molecular weight - molecular mass of compounds measured in daltons

  • Topological polar surface area - measured as an estimate of polar surface area of a molecule (i.e. the surface sum over polar atoms in a molecule), with units in angstrom squared (Å2)

  • Complexity - complexity rating for compounds, based on Bertz/Hendrickson/Ihlenfeldt formula as a rough estimation of how complex a compound was structurally

  • Partition coefficients (xlogp) - predicted octanol-water partition coefficient as a measure of the hydrophilicity or hydrophobicity of a molecule

  • Heavy atom count - number of heavy atoms e.g. non-hydrogen atoms in the compound

  • Hydrogen bond donor count - number of hydrogen bond donors in the compound

  • Hydrogen bond acceptor count - number of hydrogen bond acceptors in the compound

  • Rotatable bond count - defined as any single-order non-ring bond, where atoms on either side of the bond were in turn bound to non-terminal heavy atoms (e.g. non-hydrogen). Rotation around the bond axis would change overall molecule shape and generate conformers which could be distinguished by standard spectroscopic methods

  • Exact mass - exact mass of an isotopic species, obtained by summing masses of individual isotopes of the molecule

  • Monoisotopic mass - sum of the masses of atoms in a molecule, using unbound, ground-state, rest mass of principal (or most abundant) isotope for each element instead of isotopic average mass

  • Formal charge - the difference between the number of valence electrons of each atom, and the number of electrons the atom was associated with, assumed any shared electrons were equally shared between the two bonded atoms

  • Covalently-bonded unit count - a group of atoms connected by covalent bonds, ignoring other bond types (or a single atom without covalent bonds), representing number of such units in the compound

  • Isotope atom count - number of isotopes that were not most abundant for the corresponding chemical elements. Isotopes were variants of a chemical element that differed in neutron number

  • Defined atom stereocenter count - atom stereocenter (or chiral center) was where an atom was attached to 4 different types of atoms or groups of atoms in a tetrahedral arrangement. It could either be (R)- or (S)- configurations. Some of the compounds e.g. racemic mixtures, could have undefined atom stereocenter, where (R/S)-config was not specifically defined. Defined atom stereocenter count was the number of atom stereocenters where configurations were specifically defined

  • Undefined atoms stereocenter count - this was the undefined version of the atoms stereocenter count

  • Defined bond stereocenter count - bond stereocenter (or non-rotatable bond) was where two atoms could have different arrangement e.g. in cis- & trans- forms of butene around its double bond. Some compounds could have an undefined bond stereocenter (stereochemistry not specifically defined). Defined bond stereocenter count was the number of bond stereocenters where configurations were specifically defined.

  • Undefined bond stereocenter count - this was the undefined version of the bond stereocenter count


Convert data type for selected columns
# Convert data type - only for partition coefficients column (rest were okay)
pc_cov = pc_cov.with_column((pl.col("Partition coefficients")).cast(pl.Float64, strict = False))
pc_cov.head()
shape: (5, 38)
cid Compound name Synonyms Molecular weight Molecular formula Polar surface area Complexity Partition coefficients Heavy atom count Hydrogen bond donor count Hydrogen bond acceptor count Rotatable bond count inchi isosmiles canonicalsmiles inchikey iupacname Exact mass Monoisotopic mass Formal charge Covalently-bonded unit count Isotope atom count Total atom stereocenter count Defined atom stereocenter count Undefined atoms stereocenter count Total bond stereocenter count Defined bond stereocenter count Undefined bond stereocenter count pclidcnt gpidcnt MeSH headings annothits annothitcnt aids cidcdate sidsrcname depcatg annotation
i64 str str f64 str f64 f64 f64 i64 i64 i64 i64 str str str str str f64 f64 i64 i64 i64 i64 i64 i64 i64 i64 i64 i64 i64 str str i64 str i64 str str str
5280453 "Calcitriol" "calcitriol|322... 416.6 "C27H44O3" 60.7 688.0 5.1 30 3 3 6 "InChI=1S/C27H4... "C[C@H](CCCC(C)... "CC(CCCC(C)(C)O... "GMRQFYUYWCNGIN... "(1R,3S,5Z)-5-[... 416.329 416.329 0 1 0 6 6 0 2 2 0 22311 46029 "Calcitriol" "Biological Tes... 12 "485|631|731|78... 20040916 "A2B Chem|AA BL... "Chemical Vendo... "COVID-19, COVI...
9962735 "Ubiquinol" "ubiquinol|992-... 865.4 "C59H92O4" 58.9 1600.0 20.2 63 2 4 31 "InChI=1S/C59H9... "CC1=C(C(=C(C(=... "CC1=C(C(=C(C(=... "QNTNKSLOFHEFPK... "2-[(2E,6E,10E,... 864.7 864.7 0 1 0 0 0 0 9 9 0 2732 21358 "NULL" "Chemical and P... 7 "NULL" 20061025 "001Chemical|A2... "Chemical Vendo... "COVID-19, COVI...
5961 "Glutamine" "L-glutamine|gl... 146.14 "C5H10N2O3" 106.0 146.0 -3.1 10 3 4 4 "InChI=1S/C5H10... "C(CC(=O)N)[C@@... "C(CC(=O)N)C(C(... "ZDXPYRJPNDTMRX... "(2S)-2,5-diami... 146.069 146.069 0 1 0 1 1 0 0 0 0 88218 399 "Glutamine" "Biological Tes... 12 "422|429|436|54... 20040916 "001Chemical|3B... "Chemical Vendo... "COVID-19, COVI...
2244 "Aspirin" "aspirin|ACETYL... 180.16 "C9H8O4" 63.6 212.0 1.2 13 1 4 3 "InChI=1S/C9H8O... "CC(=O)OC1=CC=C... "CC(=O)OC1=CC=C... "BSYNRYMUTXBXSQ... "2-acetyloxyben... 180.042 180.042 0 1 0 0 0 0 0 0 0 127012 364455 "Aspirin" "Biological Tes... 12 "1|3|9|15|19|21... 20040916 "001Chemical|3B... "Chemical Vendo... "COVID-19, COVI...
457 "1-Methylnicoti... "1-methylnicoti... 137.16 "C7H9N2O+" 47.0 136.0 -0.1 10 1 1 1 "InChI=1S/C7H8N... "C[N+]1=CC=CC(=... "C[N+]1=CC=CC(=... "LDHMAVIPBRSVRG... "1-methylpyridi... 137.071 137.071 1 1 0 0 0 0 0 0 0 310 674 "NULL" "Biological Tes... 8 "61001|61002|14... 20040916 "001Chemical|3B... "Chemical Vendo... "COVID-19, COVI...


Select columns for data visualisations

The idea was really only keeping all the numerical columns for some data visualisations later. So I’ve dropped all the other columns in texts or of the string types.

# Drop unused columns in preparation for data visualisations
pc_cov = pc_cov.drop([
    "cid", 
    "Synonyms",
    "Molecular formula",
    "inchi",
    "isosmiles",
    "canonicalsmiles",
    "inchikey",
    "iupacname",
    "pclidcnt",
    "gpidcnt",
    "MeSH headings",
    "annothits",
    "annothitcnt",
    "aids",
    "cidcdate",
    "sidsrcname",
    "depcatg",
    "annotation"
])

pc_cov.head()
shape: (5, 20)
Compound name Molecular weight Polar surface area Complexity Partition coefficients Heavy atom count Hydrogen bond donor count Hydrogen bond acceptor count Rotatable bond count Exact mass Monoisotopic mass Formal charge Covalently-bonded unit count Isotope atom count Total atom stereocenter count Defined atom stereocenter count Undefined atoms stereocenter count Total bond stereocenter count Defined bond stereocenter count Undefined bond stereocenter count
str f64 f64 f64 f64 i64 i64 i64 i64 f64 f64 i64 i64 i64 i64 i64 i64 i64 i64 i64
"Calcitriol" 416.6 60.7 688.0 5.1 30 3 3 6 416.329 416.329 0 1 0 6 6 0 2 2 0
"Ubiquinol" 865.4 58.9 1600.0 20.2 63 2 4 31 864.7 864.7 0 1 0 0 0 0 9 9 0
"Glutamine" 146.14 106.0 146.0 -3.1 10 3 4 4 146.069 146.069 0 1 0 1 1 0 0 0 0
"Aspirin" 180.16 63.6 212.0 1.2 13 1 4 3 180.042 180.042 0 1 0 0 0 0 0 0 0
"1-Methylnicoti... 137.16 47.0 136.0 -0.1 10 1 1 1 137.071 137.071 1 1 0 0 0 0 0 0 0


Quick summary statistics of columns
# Overall descriptive statistics of kept columns
pc_cov.describe()
shape: (7, 21)
describe Compound name Molecular weight Polar surface area Complexity Partition coefficients Heavy atom count Hydrogen bond donor count Hydrogen bond acceptor count Rotatable bond count Exact mass Monoisotopic mass Formal charge Covalently-bonded unit count Isotope atom count Total atom stereocenter count Defined atom stereocenter count Undefined atoms stereocenter count Total bond stereocenter count Defined bond stereocenter count Undefined bond stereocenter count
str str f64 f64 f64 f64 f64 f64 f64 f64 f64 f64 f64 f64 f64 f64 f64 f64 f64 f64 f64
"count" "631" 631.0 631.0 631.0 631.0 631.0 631.0 631.0 631.0 631.0 631.0 631.0 631.0 631.0 631.0 631.0 631.0 631.0 631.0 631.0
"null_count" "0" 0.0 0.0 0.0 173.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
"mean" null 549.539675 163.915368 864.755626 2.25917 37.770206 4.066561 9.210777 9.518225 549.095022 549.06013 -0.004754 1.578447 0.006339 4.017433 3.551506 0.465927 0.381933 0.343899 0.038035
"std" null 455.236826 192.256415 1000.220379 3.926459 31.821967 6.348004 8.694184 15.393131 455.064211 454.958033 0.358537 1.610416 0.079429 6.128363 5.787792 2.364089 1.181171 1.107245 0.363159
"min" "(+)-Mefloquine... 103.1 0.0 0.0 -24.0 1.0 0.0 0.0 0.0 103.04 103.04 -6.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
"max" "sodium;8-amino... 4114.0 1650.0 9590.0 20.2 291.0 57.0 65.0 151.0 4112.12 4111.12 2.0 21.0 1.0 39.0 39.0 31.0 11.0 11.0 7.0
"median" null 435.9 110.0 635.0 2.5 30.0 3.0 7.0 6.0 435.227 435.227 0.0 1.0 0.0 1.0 1.0 0.0 0.0 0.0 0.0


Conditional assignments in Polars

The longer I’ve used Polars, the more I like its coding styles of chaining a string of different code functions together to manipulate dataframes in one go. This usually might mean that we could avoid writing some repeated loop functions to achieve the same results. In the example below, I’d like to show how to chain “when-then-otherwise” expressions by using Polars.


Chaining when-then-otherwise expressions - creating groups in data

I had the idea of separating all data into 3 different ranges of partition coefficients, so that this could be shown visually in plots. One of the possible ways (other than writing a loop function), or really the long way, to do this might be like the code shown below:

```{python}
part_coef_1 = pc_cov.filter(pl.col("Partition_coef") <= -10)
part_coef_2 = pc_cov.filter((pl.col("Partition_coef") >= -11) & (pl.col("Partition_coef") <= 5))
part_coef_3 = pc_cov.filter(pl.col("Partition_coef") >= 6)
```

A shorter and probably more elegant way was to use the “when-then-otherwise” expression in Polars for conditional assignments (the following code snippet was adapted with thanks to the author of Polars, Ritchie Vink and also the good old Stack Overflow):

pc_cov = pc_cov.with_column(
    pl.when((pl.col("Partition coefficients") <= -10))
    .then("Smaller than -10")
    .when((pl.col("Partition coefficients") >= -11) & (pl.col("Partition coefficients") <= 5))
    .then("Between -11 and 5")
    .otherwise("Larger than 6")
    .alias("Part_coef_group")
)

pc_cov.head(10)

# a new column would be added to the end of the dataframe 
# with a new column name, "Part_coef_group" 
# (scroll to the very right to see the added column)
shape: (10, 21)
Compound name Molecular weight Polar surface area Complexity Partition coefficients Heavy atom count Hydrogen bond donor count Hydrogen bond acceptor count Rotatable bond count Exact mass Monoisotopic mass Formal charge Covalently-bonded unit count Isotope atom count Total atom stereocenter count Defined atom stereocenter count Undefined atoms stereocenter count Total bond stereocenter count Defined bond stereocenter count Undefined bond stereocenter count Part_coef_group
str f64 f64 f64 f64 i64 i64 i64 i64 f64 f64 i64 i64 i64 i64 i64 i64 i64 i64 i64 str
"Calcitriol" 416.6 60.7 688.0 5.1 30 3 3 6 416.329 416.329 0 1 0 6 6 0 2 2 0 "Larger than 6"
"Ubiquinol" 865.4 58.9 1600.0 20.2 63 2 4 31 864.7 864.7 0 1 0 0 0 0 9 9 0 "Larger than 6"
"Glutamine" 146.14 106.0 146.0 -3.1 10 3 4 4 146.069 146.069 0 1 0 1 1 0 0 0 0 "Between -11 an...
"Aspirin" 180.16 63.6 212.0 1.2 13 1 4 3 180.042 180.042 0 1 0 0 0 0 0 0 0 "Between -11 an...
"1-Methylnicoti... 137.16 47.0 136.0 -0.1 10 1 1 1 137.071 137.071 1 1 0 0 0 0 0 0 0 "Between -11 an...
"Losartan" 422.9 92.5 520.0 4.3 30 2 5 8 422.162 422.162 0 1 0 0 0 0 0 0 0 "Between -11 an...
"Vitamin E" 430.7 29.5 503.0 10.7 31 1 2 12 430.381 430.381 0 1 0 3 3 0 0 0 0 "Larger than 6"
"Nicotinamide" 122.12 56.0 114.0 -0.4 9 1 2 1 122.048 122.048 0 1 0 0 0 0 0 0 0 "Between -11 an...
"Adenosine" 267.24 140.0 335.0 -1.1 19 4 8 2 267.097 267.097 0 1 0 4 4 0 0 0 0 "Between -11 an...
"Inosine" 268.23 129.0 405.0 -1.3 19 4 7 2 268.081 268.081 0 1 0 4 4 0 0 0 0 "Between -11 an...


Import Plotly

Time for some data vizzes - importing Plotly first.

import plotly.express as px


Some examples of data visualisations

Below were some of the examples of building plots by using Plotly.

Partition coefficients vs. Molecular weights
fig = px.scatter(x = pc_cov["Partition coefficients"], 
                 y = pc_cov["Molecular weight"], 
                 hover_name = pc_cov["Compound name"],
                 color = pc_cov["Part_coef_group"],
                 width = 800, 
                 height = 400,
                 title = "Partition coefficients vs. molecular weights for compounds used in COVID-19 clinical trials")

fig.update_layout(
    title = dict(
        font = dict(
            size = 15)),
    title_x = 0.5,
    margin = dict(
        l = 20, r = 20, t = 40, b = 3),
    xaxis = dict(
        tickfont = dict(size = 9), 
        title = "Partition coefficients"
    ),
    yaxis = dict(
        tickfont = dict(size = 9), 
        title = "Molecular weights"
    ),
    legend = dict(
        font = dict(
            size = 9)))

fig.show()


Molecular weights vs. Complexity
fig = px.scatter(x = pc_cov["Molecular weight"], 
                 y = pc_cov["Complexity"], 
                 hover_name = pc_cov["Compound name"],
                 #color = pc_cov["Part_coef_group"],
                 width = 800, 
                 height = 400,
                 title = "Molecular weights vs. complexity for compounds used in COVID-19 clinical trials")

fig.update_layout(
    title = dict(
        font = dict(
            size = 15)),
    title_x = 0.5,
    margin = dict(
        l = 20, r = 20, t = 40, b = 3),
    xaxis = dict(
        tickfont = dict(size = 9), 
        title = "Molecular weights"
    ),
    yaxis = dict(
        tickfont = dict(size = 9), 
        title = "Complexity"
    ),
    legend = dict(
        font = dict(
            size = 9)))

fig.show()


Export prepared dataset

Two of the possible options to export the dataset for use in a Shiny app could be:

  1. Convert Polars dataframe into a Pandas dataframe, so that it could be imported into the app for use (Polars not directly supported in Shiny for Python yet, but we could use its to_pandas() function to coerce an object e.g. a dataframe to be converted into a Pandas dataframe).

  2. Another option was to save Polars dataframe as .csv file, then read in this file in the app.py script by using Pandas (which was the method I used for this particular app)

```{python}
# --If preferring to use Pandas--
# Convert Polars df into a Pandas df if needed
df_name = df_name.to_pandas()

# Convert the Pandas df into a csv file using Pandas 
df_name.to_csv("csv_file_name.csv", sep = ",")

# --If preferring to use Polars--
# Simply write a Polars dataframe into a .csv file
df_name.write_csv("csv_file_name.csv", separator = ",")
```
Back to top ]]> Python Shiny Polars Plotly PubChem Cheminformatics https://jhylin.github.io/Data_in_life_blog/posts/13_Shiny_app_python/ShinyAppPy_PC_Cov19_data_prep.html Sun, 07 May 2023 12:00:00 GMT Shiny app in R Jennifer HY Lin https://jhylin.github.io/Data_in_life_blog/posts/12_Shiny_app_chembl/ShinyAppChembl.html
Back story

It has been a long while since I’ve used R in my projects. Last year during the time when I bravely embraced the role of rotating curator for @WeAreRLadies on Twitter around end of October, I did mention that I wanted to learn Shiny. I haven’t forgotten about this actually. So as promised, here it is, my first Shiny app in R, which is really a very simple interactive web app about small molecules from ChEMBL database. The URL to reach this app, ShinyAppChembl, is at: https://jhylin.shinyapps.io/ShinyAppChembl/. It shows a selected set of physicochemical properties for the curated small molecules in different max phases in boxplot formats. Note: it may take a few minutes to load the plot when first opening the app.


The process

Since I haven’t been using a lot of R lately, I just wanted to document how I approached this Shiny app framework, as part of my self-directed learning for R that started around mid-2022. The first place I went to was not Google’s Bard or OpenAI’s ChatGPT, as I was trying to preempt a scenario where if both of these options were temporarily down, what would I do to learn a new tool. So I visited the Shiny website first, and literally started from the “Get Started” section there, then tried to read through the lessons provided. I gathered a quick overview about the core components within a Shiny app, which were the user interface, server logic and the call to run or create app in the end, and thought to get started from there.

One of the most helpful online books called, “Mastering Shiny” had clarified a few coding issues for me. The reactivity section in the book was very useful as well to help with understanding the interactivity concept in the app. The best and also the hardest thing at this stage after reading some of the information was to actually start coding for the app in RStudio IDE, which I did soon after.


Trials-and-errors

Initially, I’ve noticed in the gallery section from the Shiny website that some of the basic Shiny apps had plots produced with R code using S3 method - the type with class ‘formula’, such as boxplot(formula e.g. y ~ group, data, and so on). So I started with this first and ended up with a draft version shown below:

First draft app using S3 method in R code to plot boxplots - screenshot taken by author


I then tried the ggplot2 version, which I preferred to use. However, I kept on hitting a roadblock repeatedly (as shown in the image below):

Second draft app using ggplot2 boxplot format - screenshot taken by author

I ended up working through this issue of not being able to display the boxplots properly over at least two days, where I tried to figure out how to change the code so that the boxplots would appear as the output in the app. I actually wrote a plot function code (as shown below) before working on the app.R file, in order to trial plotting the boxplots, making sure that the code worked before using it in the app.R file.

```{r}
dfBoxplot <- function(var) {
  label <- rlang::englue("{{var}} vs. Max Phases of small molecules")
  
  chembl %>% 
    select(`Max Phase`, {{ var }}) %>% 
    ggplot(aes(x = `Max Phase`, y = {{ var }})) +
    geom_boxplot(aes(group = cut_width(`Max Phase`, 0.25), 
                     colour = `Max Phase`), 
                 outlier.alpha = 0.2) +
    labs(title = label)
}
```

Once I made sure this code worked, I transplanted the code into the server section of the app.R file, however it wasn’t that simple obviously. Through the process of more trials-and-errors, I managed to figure out the code for the plot output in the final version, which was not the same as the function code above, but more like this.

```{r}
  output$BPlot <- renderPlot({ 
    
    ggplot(chembl, aes(`Max Phase`, .data[[input$variable]])) +
      geom_boxplot(aes(group = cut_width(`Max Phase`, 0.25), 
                       colour = `Max Phase`), outlier.alpha = 0.2) +
      labs(title = "Distributions of physicochemical properties against max phases",
           caption = "(based on ChEMBL database version 31)") +
      theme_minimal()
    
    }, res = 96) %>% bindCache(chembl$`Max Phase`, input$variable)
```

I then read about the section on “Tidy evaluation” in the “Mastering Shiny” book, which had thoroughly described the problems I’ve encountered (and which I wished I had actually read this section before and not after hitting the roadblock…). So I’d highly recommend new users to read this section and also the rest of the book if Shiny’s also new to you.


Final app

The final app now looks like this:

Screenshot taken by author


App deployment

After I got the app working, I looked into where I could deploy the app, since my main goal was to learn and share my work. At first, I went to the Shiny section on the Quarto website to see if it was possible to deploy the app in Quarto. However, after reading through several questions and answers in relation to Shiny apps and Quarto website, it was obvious that it was still not possible yet to deploy the app in an interactive way on Quarto websites (but it was mentioned in Posit community that this was being looked into, so I’m looking forward to the day when we can do exactly that in the future). This means that currently, there will only be an app image showing up in a Quarto document at most. I ended up choosing shinyapp.io to deploy my first Shiny app for now.


About the boxplots

Since the main goal of this post is more on the process of producing a simple Shiny app for a new comer, I won’t go into the fine details to describe how these boxplots differ between different max phases. Also as a side note, I’m aware that some experts in data visualisations might not really like boxplots in general, but for my case, I’ve got molecules in different max phases where a boxplot is presented for each max phase lining up next to each other. Therefore, in a way, some relative comparisons or differences could be drawn visually in the first glance, although other graph types such as density plots or heat maps might be better options.

I’ll focus on the “QED Weighted” variable here, as it’s a physicochemical property that has combined several molecular features together as a score (please refer to this post - section: “Some exploratory data analysis” for details about this QED weighted score). For all the boxplots shown when “QED Weighted” is selected from the drop-down box, max phase 4 molecules obviously have higher QED weighted scores in general than all of the other max phases. This is especially clear when comparing the medians between them, with max phase 4 small molecules having a median QED weighted score of more than 0.5, and the rest of the other max phases had 0.5 or below. The higher the QED weighted scores, the more druglike the molecules will be, and for max phase 4 molecules, they are mostly prescription medicines that have already reached approval and are already being widely prescribed. So this makes sense as this is being reflected in the boxplots for these ChEMBL small molecules.


Final words

Finally, I’m quite pleasantly surprised that there is also a Shiny in Python version, which has a Shinylive option to deploy Shiny app interactively in GitHub Gist and so on… I’ll most likely need to read further into this and make this as my next project. This is also a nice break from my recent machine learning projects, which I’ll try to return to once I’ve had enough fun with Shiny!

Thanks for reading.

Back to top ]]> R Shiny ChEMBL database Cheminformatics https://jhylin.github.io/Data_in_life_blog/posts/12_Shiny_app_chembl/ShinyAppChembl.html Thu, 06 Apr 2023 12:00:00 GMT Pills dataset - Part 3 Jennifer HY Lin https://jhylin.github.io/Data_in_life_blog/posts/09_Pills/Rust_evcxr_polars_plotly_final.html
Background

The aim of this final part (part 3) for the pills dataset was really for me to start using Rust in a beginner-friendly way. Overall, this trilogy (parts 1 - 3) for the pills dataset formulated an overview of how to use Polars in Python (mainly), Pandas in Python (smaller section) and Polars in Rust (even little less as this was new to me) with Plotly. Over time, I’ve been finding myself learning more optimally by doing and applying, rather than just reading and thinking, so I’ve got myself started in this very new programming language, Rust, to get some familiarities. I anticipated that I would still work with Python and R mainly in the near future, so that I’m not diverting too much and would be at least proficient in at least one programming language.

My very initial idea was to integrate Rust-Polars, Plotly in Rust (Plotly.rs) and Jupyter-Evcxr together, and see if I could get a simple data visualisation out of a small dataset. Although the idea sounded simple enough, I was actually quite stuck at the step of importing one of the columns as x-axis variables in Rust-Polars to Plotly.rs. I figured it might possibly be due to my very lack-of-knowledge and lack-of-familiarities with Rust (I do need to continue reading the Rust programming language book), Polars (I’m better with Python-Polars actually), Plotly.rs and also Evcxr. Another possibility could be that Plotly.rs mainly had ndarray support, and Polars was not mentioned explicitly in Plotly.rs so my guess was that these two might not flow very well together. Also, Polars itself was constantly evolving and growing as well.

So I’ve decided to leave things as how it would be for now, before I delayed this post any further. If I happened to figure out how to do this in the future, then I’ll come back to update this last part of the project. While I was tackling this little issue mentioned above, somehow I’ve managed to deconstruct Polars dataframe in Rust in Evcxr. So I’ll show a little bit about it below. One slightly good news that came out from all of this, was that I’ve managed to import the other column as y-axis variables, which contained numbers, without problems. I’ve also figured out the Rust code to convert Series/ChunkedArray to vectors in Rust IDEs (e.g. VS Code, and quite a few others). So I did learn a few things while completing this post, and hoped I could expand further on this later.

Note: I’ve published all Rust code as print-only in Quarto markdown file, since it’s not possible to run them in RStudio IDE (Rust was not supported). So all Rust code were originally run on Jupyter Lab in MacOS, with code outputs being captured as screenshots, which were shown as photos in this post. Here’s the link to the .ipynb file in the GitHub repository for this portfolio website (or alternatively, you could access it from the GitHub icon link at the top of the web page), in case anyone wanted to see the full .ipynb version.


Import dependencies

These dependencies were known as crates in the world of Rust. I’d also like to think of them as libraries or packages we would install or import in Python and R. So this step was necessary before I even started anything decent in Rust. Similar things would also apply to Rust IDEs as well since I’ve played a little bit in VS Code previously.

```{rust}
// Set up required dependencies
:dep ndarray = "0.15.6"
```
```{rust}
:dep plotly = { version = ">=0.8.0", features = ["plotly_ndarray"]}
```
```{rust}
// May take a few minutes to load polars crate (might depend on your machine specs)
:dep polars = { version = ">=0.26.0", features = ["lazy", "csv-file", "strings", "dtype-duration", "dtype-categorical", "concat_str", "rank", "lazy_regex", "ndarray"]}
```
```{rust}
:dep itertools = {version = "0.9.0"}
```


Import external crates
```{rust}
// Import external crates needed
extern crate ndarray;
extern crate plotly;
extern crate polars;
```


Specify imports or modules required
```{rust}
use ndarray::prelude::*;
use polars::prelude::*;
use plotly::common::{
    ColorScale, ColorScalePalette, DashType, Fill, Font, Line, LineShape, Marker, Mode, Title,
};
use plotly::layout::{Axis, BarMode, Layout, Legend, TicksDirection};
use plotly::{Plot, Scatter, Bar};
use itertools::Itertools;
```


Reading csv file

Here, I’ve imported the .csv file saved from part 2 of the project.

```{rust}
// Reading .csv file
let df = CsvReader::from_path("ace_para_count.csv").unwrap().finish().unwrap();
```
```{rust}
df
```

Photo by author


Converting columns into ndarrays

I’ve tested plotting in Plotly.rs after a few trials and errors at the beginning, but luckily I’ve spotted the ndarray support from the Plotly.rs book soon enough to figure out that I could convert the “count” column into a ndarray first, which was shown in the code below.

```{rust}
// Switch Polars dataframe into 2D array
// Ensure "ndarray" was added as one of the features for polars under dependencies

/*Example from Polars documentation:
let df = DataFrame::new(vec![a, b]).unwrap();
let ndarray = df.to_ndarray::<Float64Type>().unwrap();
println!("{:?}", ndarray);
*/

//Note: ndarray for numbers only, not strings, so only "count" column was converted
let ndarray = df.to_ndarray::<Float64Type>().unwrap();
println!("{:?}", ndarray);
```

Photo by author


Deconstructing Polars dataframe in Rust

Because “to_ndarray” was only for numerics and not strings, I ran into a problem trying to figure out how to best import this other “Colour” column into Plotly.rs. This led to my little convoluted journey to work with Polars dataframe in Rust, trying to see if I could convert the “Colour” column into a vector (which might not be the best way to do it, but as part of my Rust learning, I went for it anyway). I’ve subsequently tried plotting the “count” column in ndarray as a vector with success, based on the reference from Plotly.rs book that variables for x or y-axis could be placed into a vector by using a vector macro. Eventually, I didn’t quite achieve my goal but I’ve managed to break down or convert the Polars dataframe into different formats.

```{rust}
// Select specific column or series by position
let Colours = df[0].clone();

//Alternative way to select specific column or series by name
//let u = df.select_series(&["Colour"]);
```
```{rust}
Colours
```

Photo by author

There was a mention of storing series (column) in a vec (as series vector, not vector for strings) in Polars’ documentation, which I’ve tried to plot in Plotly.rs, but it unfortunately failed to work. One of my guesses could be due to the data type used for vector, as Rust was a very type-specific programming language, which also brought its well-known memory safety and other benefits in the long run. My immediate thought was that it probably needed to be a vector for strings, not series, which might make it work. Then I was searching on StackOverflow for similar questions and answers, then I found something related to what I wanted to do from Polars documentations as shown below.

```{rust}
// Adapted from: https://docs.rs/polars/latest/polars/docs/eager/index.html#series 
// Extracting data: 
// To be able to extract data out of Series, 
// either by iterating over them or converting them to other datatypes like a Vec<T>, 
// we first need to downcast them to a ChunkedArray<T>. 
// This is needed because we don't know the data type that is held by the Series.

/*use polars::prelude::*; 
  use polars::df;

  fn extract_data() -> PolarsResult<()> { 
  let df = df! [ "a" => [None, Some(1.0f32), Some(2.0)], "str" => ["foo", "bar", "ham"]]?;

// first extract ChunkedArray to get the inner type.

  let ca = df.column("a")?.f32()?;

// Then convert to vec

  let _to_vec: Vec<Option<f32>> = Vec::from(ca);

// We can also do this with iterators

  let ca = df.column("str")?.utf8()?; 
  let _to_vec: Vec<Option<&str>> = ca.into_iter().collect(); 
  let _to_vec_no_options: Vec<&str> = ca.into_no_null_iter().collect();

  Ok(())

}*/
```

Initially, I trialled the iterator function first.

```{rust}
// Print out items in column by applying an iterator to it
println!("{}", &Colours.iter().format("\n"));
```

Photo by author

Then, it took me quite a long time to just downcast Series into ChunkedArray, but somehow I’ve managed to figure out the code myself below. One of the likely reasons was due to my choice of using Evcxr, which required Rust code in slightly different formats than the ones in Rust IDEs (although almost the same).

```{rust}
// Somehow worked out how to convert series to chunkedarray by accident!
println!("{:?}", Colours.utf8().unwrap());
```

Photo by author

Then I moved onto trying to figure out how to convert or place a ChunkedArray into a vector, with the closest answer shown below. However, bear in mind that these Rust code were for Rust IDEs, and not for Evcxr, so this added slightly more complexities to what I was trying to do (perhaps I should just stick with Rust IDEs in the future…).

```{rust}
//Adpated from StackOverflow - How to get a Vec from polars Series or ChunkedArray?
//You can collect the values into a Vec.

/*use polars::prelude::*;

fn main() -> Result<()> { let s = Series::new("a", 0..10i32);

    let as_vec: Vec<Option<i32>> = s.i32()?.into_iter().collect();

    //if we are certain we don't have missing values
    let as_vec: Vec<i32> = s.i32()?.into_no_null_iter().collect();
    Ok(())

}*/
```

I also found another way to iterate the ChunkedArray.

```{rust}
//fn iter_forward(ca: &Float32Chunked) { ca.into_iter().for_each(|opt_v| println!("{:?}", opt_v)) } 
```

I then tested another iterator method, and came up with another line of code as shown below, which listed the colours in the “Colour” column.

```{rust}
Colours.utf8().unwrap().into_iter().for_each(|array|println!("{:?}", array));
```

Photo by author

I also found out, randomly, how to slice strings for Series in Polars. By changing the number of letters to slice through in Some(), the strings or words would vary length accordingly. Here, I’ve used “15” so it covered all the colours (note: the longest would be 12 characters for combination colours).

```{rust}
// Another method to use if needing to slice strings
let x = Colours.utf8().unwrap().str_slice(0, Some(15));
x
```

Photo by author

Lastly, before I got too carried away, I just wanted to show the method from Polars documentations that this was the Polars’ way to select a specific column from Polars dataframe.

```{rust}
let ca = df.clone().lazy().select([cols(["Colour"])]).collect()?;
ca
```

Photo by author


Plotting Polars dataframe in Plotly.rs

For the x-axis, eventually, I reverted for manual input due to the issue mentioned in the background section. So the colours from the “Colour” column were stored in a vector set up manually, rather than coming directly from the dataframe. While searching for answers, I’ve also learnt several other tricks, although not really solving the problem, they might still be useful in the future. For the y-axis, the ndarray for the “count” column was converted into a vector first before being fed into the trace (graph module), and thankfully the plot worked nicely.

```{rust}
// MANUAL method:
// Use vec! macro to create new vectors to hold x variables (words as strings)
// Manually input the colour names (as ndarray is only for numbers)
let x = vec!["RED", "ORANGE;BROWN", "YELLOW;WHITE", "ORANGE", "WHITE", "BLUE"];

// Plot using ndarray, which is supported by Plotly.rs 
// Polars likely not supported yet
// Convert ndarray (holding counts as y variables) into vector 
let y = ndarray.column(1).to_vec();

// Use trace as a graph module,
// choose type of plots needed for x & y variables called
// Graph options e.g. Scatter, Line or Bar
let trace = Scatter::new(x, y);

// Set plot variable as mutable and initiate a plot
let mut plot = Plot::new();
// Add trace (graph) into the plot variable
plot.add_trace(trace);

// Specify the specs for plot
let layout = Layout::new()
    // Choose height of graph
    .height(500)
    // Name x-axis
    .x_axis(Axis::new().title(Title::new("Colours")))
    // Name y-axis
    .y_axis(Axis::new().title(Title::new("Count")))
    // Add title of graph
    .title(Title::new("Frequency of colours in acetaminophen (paracetamol) oral dosage forms"));

// Set the layout of the plot
plot.set_layout(layout);

// Display the plot in Jupyter Lab format 
// For Jupyter Notebook, use: plot.notebook_display();
plot.lab_display();
format!("EVCXR_BEGIN_CONTENT application/vnd.plotly.v1+json\n{}\nEVCXR_END_CONTENT", plot.to_json())
```

Photo by author


Conclusion

This last part was the hardest for me to execute out of all 3 parts (it likely took me a good whole week to figure out deconstructing Polars dataframe and trying to work with vectors), as Rust was completely new to me. At one point I thought about jumping back to Python, but I persisted and although I didn’t quite solve the string importation issue, I was somehow happy that I was at least able to see how this programming language could be applied in Polars dataframe library. I also got a taste of using Rust in data visualisations. All I wanted to show was that there were a variety of data tools to use, and knowing your tools of trade would be the most critical when working on different data projects as certain tools would only work the best for certain tasks and scenarios. This warm-up lesson in Rust was quite interesting and I might continue either in VS Code or Evcxr depending on my next topic of interest.


References

Rust programming language book: https://doc.rust-lang.org/book/title-page.html

Polars crate documentations: https://docs.rs/polars/latest/polars/

Plotly.rs GitHub repository: https://github.com/igiagkiozis/plotly (link to the Plotly.rs book can be found in “Converting columns into ndarrays” section)

Evcxr GitHub repository: https://github.com/google/evcxr

Back to top ]]> Data analytics projects Pills dataset series Rust Polars Plotly Evcxr Jupyter https://jhylin.github.io/Data_in_life_blog/posts/09_Pills/Rust_evcxr_polars_plotly_final.html Mon, 13 Feb 2023 11:00:00 GMT Pills dataset - Part 2 Jennifer HY Lin https://jhylin.github.io/Data_in_life_blog/posts/09_Pills/Rust_polars_pills_df.html
Quick overview

Part 2 of this project aimed to look at the pills data up close, particularly into the types of dosage forms, colours, shapes and inactive excipients used in oral medications. Plotly was used as the main data visualisation library, which was followed by some text cleaning for a particularly busy column in the dataset. This was then completed with a section in the end to generate a small dataframe, preparing for a simple data visualisation in Rust-Evcxr for the final part of this project (part 3).


Photo by Myriam Zilles on Unsplash


Import libraries and pills dataset
import polars as pl
import plotly.express as px

The pills.csv file saved from part 1 was imported as shown below.

df = pl.read_csv("pills.csv")
df
shape: (83925, 5)
splshape_text splcolor_text spl_strength spl_inactive_ing DosageForm
str str str str str
"CAPSULE" "PINK" "TEMAZEPAM 15 m... "SILICON DIOXID... "CAPSULE"
"ROUND" "ORANGE" "IBUPROFEN 200 ... "SILICON DIOXID... "TABLET, FILM C...
"PENTAGON (5 SI... "GREEN" "DEXAMETHASONE ... "ANHYDROUS LACT... "TABLET"
"ROUND" "WHITE" "Nickel Sulfate... null "TABLET"
"ROUND" "WHITE" "CLONAZEPAM 0.2... "SORBITOL;ASPAR... "TABLET, ORALLY...
"ROUND" "WHITE" "SILDENAFIL CIT... "ANHYDROUS DIBA... "TABLET, FILM C...
"OVAL" "YELLOW" "RISPERIDONE 3 ... "LACTOSE MONOHY... "TABLET, FILM C...
"CAPSULE" "BLUE" "IBUPROFEN 200 ... "FD&C BLUE NO. ... "CAPSULE, LIQUI...
"ROUND" "WHITE" "Iloperidone 12... "silicon dioxid... "TABLET"
"CAPSULE" "YELLOW;WHITE" "FENOPROFEN CAL... "CROSPOVIDONE;M... "CAPSULE"
"ROUND" "YELLOW" "BUTALBITAL 50 ... "STARCH, CORN;C... "TABLET"
"ROUND" "WHITE" "ESTRADIOL 0.5 ... "COPOVIDONE K25... "TABLET"
... ... ... ... ...
"ROUND" "WHITE" "MEMANTINE HYDR... "SILICON DIOXID... "TABLET"
"CAPSULE" "ORANGE" "ACETAMINOPHEN ... "BUTYLATED HYDR... "CAPSULE, LIQUI...
"ROUND" "WHITE" "LAMOTRIGINE 25... "MAGNESIUM CARB... "TABLET, CHEWAB...
"OVAL" "BLUE" "ACETAMINOPHEN ... "ACESULFAME POT... "TABLET, COATED...
"OVAL" "WHITE" "AZITHROMYCIN D... "CROSCARMELLOSE... "TABLET, FILM C...
"OVAL" "BLUE" "IBUPROFEN 200 ... "FD&C BLUE NO. ... "CAPSULE, LIQUI...
"OVAL" "WHITE" "CETIRIZINE HYD... "STARCH, CORN;H... "TABLET"
"OVAL" "BROWN" "OMEPRAZOLE 20 ... "CARNAUBA WAX;F... "TABLET, DELAYE...
"ROUND" "PINK;ORANGE;YE... "CALCIUM CARBON... "CITRIC ACID MO... "TABLET, CHEWAB...
"OVAL" "GREEN" "ACETAMINOPHEN ... "STARCH, CORN;D... "TABLET, FILM C...
"CAPSULE" "BLUE" "Amlodipine bes... "Cellulose, mic... "CAPSULE"
"ROUND" "ORANGE" "DARIFENACIN 15... "ANHYDROUS DIBA... "TABLET, EXTEND...


Change column names

Again, column names were changed to something easier to read.

# Rename all column names
df_new = df.rename({"splcolor_text": "Colour", 
                    "splshape_text": "Shape", 
                    "spl_strength": "Drug_strength", 
                    "spl_inactive_ing": "Inactive_excipients", 
                    "DosageForm": "Dosage_form"}
                  )
df_new
shape: (83925, 5)
Shape Colour Drug_strength Inactive_excipients Dosage_form
str str str str str
"CAPSULE" "PINK" "TEMAZEPAM 15 m... "SILICON DIOXID... "CAPSULE"
"ROUND" "ORANGE" "IBUPROFEN 200 ... "SILICON DIOXID... "TABLET, FILM C...
"PENTAGON (5 SI... "GREEN" "DEXAMETHASONE ... "ANHYDROUS LACT... "TABLET"
"ROUND" "WHITE" "Nickel Sulfate... null "TABLET"
"ROUND" "WHITE" "CLONAZEPAM 0.2... "SORBITOL;ASPAR... "TABLET, ORALLY...
"ROUND" "WHITE" "SILDENAFIL CIT... "ANHYDROUS DIBA... "TABLET, FILM C...
"OVAL" "YELLOW" "RISPERIDONE 3 ... "LACTOSE MONOHY... "TABLET, FILM C...
"CAPSULE" "BLUE" "IBUPROFEN 200 ... "FD&C BLUE NO. ... "CAPSULE, LIQUI...
"ROUND" "WHITE" "Iloperidone 12... "silicon dioxid... "TABLET"
"CAPSULE" "YELLOW;WHITE" "FENOPROFEN CAL... "CROSPOVIDONE;M... "CAPSULE"
"ROUND" "YELLOW" "BUTALBITAL 50 ... "STARCH, CORN;C... "TABLET"
"ROUND" "WHITE" "ESTRADIOL 0.5 ... "COPOVIDONE K25... "TABLET"
... ... ... ... ...
"ROUND" "WHITE" "MEMANTINE HYDR... "SILICON DIOXID... "TABLET"
"CAPSULE" "ORANGE" "ACETAMINOPHEN ... "BUTYLATED HYDR... "CAPSULE, LIQUI...
"ROUND" "WHITE" "LAMOTRIGINE 25... "MAGNESIUM CARB... "TABLET, CHEWAB...
"OVAL" "BLUE" "ACETAMINOPHEN ... "ACESULFAME POT... "TABLET, COATED...
"OVAL" "WHITE" "AZITHROMYCIN D... "CROSCARMELLOSE... "TABLET, FILM C...
"OVAL" "BLUE" "IBUPROFEN 200 ... "FD&C BLUE NO. ... "CAPSULE, LIQUI...
"OVAL" "WHITE" "CETIRIZINE HYD... "STARCH, CORN;H... "TABLET"
"OVAL" "BROWN" "OMEPRAZOLE 20 ... "CARNAUBA WAX;F... "TABLET, DELAYE...
"ROUND" "PINK;ORANGE;YE... "CALCIUM CARBON... "CITRIC ACID MO... "TABLET, CHEWAB...
"OVAL" "GREEN" "ACETAMINOPHEN ... "STARCH, CORN;D... "TABLET, FILM C...
"CAPSULE" "BLUE" "Amlodipine bes... "Cellulose, mic... "CAPSULE"
"ROUND" "ORANGE" "DARIFENACIN 15... "ANHYDROUS DIBA... "TABLET, EXTEND...


Visualising oral dosage forms & colours in pills

Grabbing only unique drugs in the dataset to minimise duplications.

df_viz = df_new.unique(subset = "Drug_strength")
df_viz
shape: (9287, 5)
Shape Colour Drug_strength Inactive_excipients Dosage_form
str str str str str
"CAPSULE" "PINK" "TEMAZEPAM 15 m... "SILICON DIOXID... "CAPSULE"
"ROUND" "ORANGE" "IBUPROFEN 200 ... "SILICON DIOXID... "TABLET, FILM C...
"PENTAGON (5 SI... "GREEN" "DEXAMETHASONE ... "ANHYDROUS LACT... "TABLET"
"ROUND" "WHITE" "Nickel Sulfate... null "TABLET"
"ROUND" "WHITE" "CLONAZEPAM 0.2... "SORBITOL;ASPAR... "TABLET, ORALLY...
"ROUND" "WHITE" "SILDENAFIL CIT... "ANHYDROUS DIBA... "TABLET, FILM C...
"OVAL" "YELLOW" "RISPERIDONE 3 ... "LACTOSE MONOHY... "TABLET, FILM C...
"ROUND" "WHITE" "Iloperidone 12... "silicon dioxid... "TABLET"
"CAPSULE" "YELLOW;WHITE" "FENOPROFEN CAL... "CROSPOVIDONE;M... "CAPSULE"
"ROUND" "YELLOW" "BUTALBITAL 50 ... "STARCH, CORN;C... "TABLET"
"ROUND" "WHITE" "ESTRADIOL 0.5 ... "COPOVIDONE K25... "TABLET"
"CAPSULE" "YELLOW" "BENZONATATE 20... "D&C YELLOW NO.... "CAPSULE"
... ... ... ... ...
"ROUND" "WHITE" "BUSULFAN 2 mg" "HYPROMELLOSES ... "TABLET, FILM C...
"OVAL" "ORANGE" "AMPHETAMINE SU... "CELLULOSE, MIC... "TABLET"
"ROUND" "GREEN" "FOLIC ACID 800... "MICROCRYSTALLI... "TABLET"
"ROUND" "WHITE" "ACONITUM NAPEL... "LACTOSE / MAGN... "TABLET"
"ROUND" "WHITE" "MATRICARIA CHA... "ACACIA;LACTOSE... "TABLET"
"OVAL" "RED" "FOLIC ACID 1 m... "CELLULOSE, MIC... "TABLET"
"ROUND" "BLUE" "Labetalol 300 ... "ANHYDROUS LACT... "TABLET, FILM C...
"CAPSULE" "YELLOW;BLUE" "POMALIDOMIDE 1... "MANNITOL;STARC... "CAPSULE"
"CAPSULE" "PURPLE" "CALCIUM CITRAT... "GELATIN;" "CAPSULE, GELAT...
"CAPSULE" "GREEN;YELLOW" "LENALIDOMIDE 1... "ANHYDROUS LACT... "CAPSULE"
"OVAL" "WHITE" "folic acid 1 m... "cellulose, mic... "TABLET"
"CAPSULE" "BLUE" "Amlodipine bes... "Cellulose, mic... "CAPSULE"

One way to avoid switching Polars dataframe to a Pandas one, which could be one of the options to plot data from Polars dataframes in Plotly, was to call the x-axis and y-axis data directly from the dataframe as shown in the code below.

# scatter plot for colours, dosage forms & drug strengths 
fig = px.scatter(x = df_viz["Colour"], 
                 y = df_viz["Dosage_form"], 
                 color = df_viz["Colour"],
                 hover_name = df_viz["Drug_strength"],
                 width = 900, 
                 height = 400,
                 title = "Oral dosage forms and colours of pills")

# Update layout of the plot
fig.update_layout(
    # Change title font size
    title = dict(
        font = dict(
            size = 15)),
    # Centre the title
    title_x = 0.5,
    # Edit margins
    margin = dict(
        l = 20, r = 20, t = 40, b = 3),
    # Change x-axis
    xaxis = dict(
        tickfont = dict(size = 9), 
        title = "Colours"
    ),
    # Change y-axis
    yaxis = dict(
        tickfont = dict(size = 9), 
        title = "Dosage forms"
    ),
    # Edit lengend font size
    legend = dict(
        font = dict(
            size = 9)))

fig.show()


White was the most common colour, especially after zooming in the plot. Capsule was very commonly used as the oral dosage form of choice in this dataset.


Visualising shapes & colours in pills
fig = px.scatter(x = df_viz["Colour"], 
                 y = df_viz["Shape"], 
                 color = df_viz["Colour"],
                 hover_name = df_viz["Drug_strength"],
                 width = 900, 
                 height = 400,
                 title = "Shapes and colours of pills")

# Update layout of the plot
fig.update_layout(
    # Change title font size
    title = dict(
        font = dict(
            size = 15)),
    # Centre the title
    title_x = 0.5,
    # Edit margins
    margin = dict(
        l = 20, r = 20, t = 40, b = 3),
    # Change x-axis
    xaxis = dict(
        tickfont = dict(size = 9), 
        title = "Colours"
    ),
    # Change y-axis
    yaxis = dict(
        tickfont = dict(size = 9), 
        title = "Shapes"
    ),
    # Edit lengend font size
    legend = dict(
        font = dict(
            size = 9)))

fig.show()