Useful functions and scripts for working with small molecules.
Optionally, create a conda environment.
conda create -y -n chemfunc python=3.10
conda activate chemfuncInstall the latest version of Chem Func using pip.
pip install chemfunc
Alternatively, clone the repository and install the local version of the package.
git clone https://github.com/swansonk14/chemfunc.git
cd chemfunc
pip install -e .
If there are version issues with the required packages, create a conda environment with specific working versions of the packages as follows.
pip install -r requirements.txt
pip install -e .Note: If you get the issue ImportError: libXrender.so.1: cannot open shared object file: No such file or directory, run conda install -c conda-forge xorg-libxrender.
Chem Func contains a variety of useful functions and scripts for working with small molecules.
Functions can be imported from the chemfunc package. For example:
from pathlib import Path
from chemfunc.sdf_to_smiles import sdf_to_smiles
sdf_to_smiles(
data_path=Path('molecules.sdf'),
save_path=Path('molecules.csv')
)Most modules can also be run as scripts from the command line using the chemfunc command along with the appropriate function name. For example:
chemfunc sdf_to_smiles \
--data_path molecules.sdf \
--save_path molecules.csvTo see a list of available scripts, run chemfunc -h.
For each script, run chemfunc <script_name> -h to see a description of the arguments for that script.
Below is a list of the contents of the package.
canonicalize_smiles.py (function, script)
Canonicalizes SMILES using RDKit canonicalization and optionally strips salts.
chemical_diversity.py (function, script)
Computes the chemical diversity of a set of molecules in terms of Tanimoto distances.
cluster_molecules.py (function, script)
Performs k-means clustering to cluster molecules based on Morgan fingerprints.
compute_property_distribution.py (function, script)
Computes one or more molecular properties for a set of molecules.
deduplicate_smiles.py (function, script)
Deduplicate a CSV files by SMILES.
filter_molecules.py (function, script)
Filters molecules to those with values in a certain range.
measure_experimental_reproducibility.py (function, script)
Measures the experimental reproducibility of two biological replicates by using one replicate to predict the other.
molecular_fingerprints.py (functions, script)
Contains functions to compute fingerprints for molecules. Parallelized for speed. The function save_fingerprints can be used as a script to compute fingerprints from a CSV file and save them as an NPZ file.
molecular_properties.py (functions)
Contains functions to compute molecular properties. Parallelized for speed.
molecular_similarities.py (functions)
Contains functions to compute similarities between molecules. Parallelized for speed.
nearest_neighbor.py (function, script)
Given a dataset of molecules, computes the nearest neighbor molecule in a second dataset using one of several similarity metrics.
plot_property_distribution.py (function, script)
Plots the distribution of molecular properties of a set of molecules.
plot_tsne.py (function, script)
Runs a t-SNE on molecular fingerprints from one or more chemical libraries.
regression_to_classification.py (function, script)
Converts regression data to classification data using given thresholds.
sample_molecules.py (function, script)
Samples molecules from a CSV file, either uniformly at random across the entire dataset or uniformly at random from each cluster within the data.
sdf_to_smiles.py (function, script)
Converts an SDF file to a CSV file with SMILES.
select_from_clusters.py (function, script)
Selects the best molecule from each cluster.
smiles_to_svg.py (function, script)
Converts a SMILES string to an SVG image of the molecule.
visualize_molecules.py(function, script)
Converts a file of SMILES to images of molecular structures.
visualize_reactions.py (function, script)
Converts a file of reaction SMARTS to images of chemical reactions.