The purpose of this repository is to collect useful scripts which mainly use RDKit. Contributions are welcome!
Some scripts may require further dependencies.
- There is a
read_input.pyscript which contains the functionread_input. It reads molecules from SMI, SDF, SDF.GZ and PKL (pickled molecules as tuples of mol and mol_title) files and STDIN (SMI and SDF formats are supported) and it returns tuples of (mol, mol_title). This is a generator and can be applied to process large collections of molecules. I advise to use this function if you do not need other data from input files. - There is
_template.pyfile which can be used as a template for new scripts. Please do not change names for input, output, ncpu and verbose arguments. This will help to make command line arguments consistent across scripts. - Add help messages to your scripts.
- Ideally scripts should be able to communicate with STDIN and STDOUT to combine them with pipes. I implemented this in
gen_stereo_rdkit.pyandgen_conf_rdkit.py. - All scripts can contain errors, so use them on your own risk. If you will find a mistake please create the issue and we will fix it. However, we constantly revise old scripts and fix errors because every found mistake is penultimate.
add_prefix - add a prefix to molecule names in SDF file.
extractsdf - extract molecule names and field values from input SDF.
extract_mol_by_name - extract molecules by name (partial name matching) to new SDF file
insert_sdf - add data from a text file as additional fields to input SDF file
remove_dupl_by_field - remove entries from SDF file having duplicated mol title or field value.
rename_mols - identify identical entries in SDF (conformers) and rename them in identical manner.
sdf_field2title - insert values of a given SDF field into molecular title, or use SMILES as titles or enumerate titles sequentially.
sdf_title2field - insert molecular title into a given SDF field
strip_blank_lines - remove empty lines in multi-line field values in input SDF.
cansmi - return canonical SMILES of input molecules.
frags2mols - save disconnected components of input molecules as individual molecules with added suffix to the name.
molchemaxon2pdb - convert molecules from the input file to separate pdb files. Conformer generation is performed by RDKit. Major tautomeric form at the given pH is generated by ChemAxon.
mols2pdb - convert input molecules from SMILES or SDF file to individual PDB files. Hydrogens will be added and a random conformer will be generated if the molecule does not have 3D coordinates.
pkl2sdf - convert PKL file to SDF file. Specifically useful for conversion of generated conformers stored in PKL format by gen_conf_rdkit.
sdf2mols - split SDF to multiple MOL files.
sdf2pkl - convert SDF to multi-conformer PKL file. Conformers are recognized by mol title and should go sequentially in input SDF.
smi2sdf - Convert SMILES to SDF including additional fields if they are named and exist in SMILES file
split_pdb - split PDB by chains and save to separate PDB files.
Manipulate with Mol objects (calc properties, generate conformers/stereoisomers, filter compounds, etc):
add_h - hydrogenize input files
calc_center_rdkit - calculate the center of coordinates of all atoms in a molecule(s).
count_undefined_stereocenters - return to STDOUT names and the number of undefined stereocenters in input molecules.
discard_compounds_rdkit - remove multi-component compounds and compounds with non-organic atoms.
draw_mols - return PNG images of input molecules.
filter_conf - filter conformers by RMS value.
filter_conf_adv - selection of representative conformers by RMS value using clustering and advanced features (e.g. preferable selection of specifically labeled conformers).
gen_conf_rdkit - generate conformers.
gen_stereo_rdkit - enumerate stereoisomers (tetrahedral and double bond).
get_mol_center - returns a geometrical center of a molecule
get_substr - filter input molecules by SMARTS, multiple SMARTS are allowed, negative matching is possible.
get_total_charge - calculate total formal charge of input MOL files.
keep_largest - keep the largest fragment by the number of heavy atoms in each compound record. If components have the same number a random one will be selected.
mirror_mols - return mirrored 3D input structure and optionally rename it. Useful for generation of enantiomers of molecules with axial/planar chirality.
murcko - return Murcko scaffolds ignoring stereoconfiguration.
physchem_calc - calculate various physicochemical properties of input molecules (MW, logP, TPSA, QED, etc).
pmapper_descriptors - calculate 3D pharmacophore descriptors with pmapper and remove rarely occurred ones. Useful for QSAR modeling.
remove_dupl_rdkit - remove duplicates by InChi keys comparison within the input file or relatively to a reference file.
rmsd_rdkit - calculate RMSD for input MOL2/PDBQT/SDF files. Automatically calculate RMSD for maximum common substructure if full atom matching was failed. Symmetry checking was implemented.
sanitize_rdkit - remove compounds with RDKit sanitization errors and add to output molecules the number of double bonds, unspecified stereocenters and total charge.
sphere_exclusion - return names of a diverse subset of input compounds
test_pains - return a list of SMILES matched PAINS.
binning - take a table with variable values and return a table with binned values according to supplied thresholds.
vina_dock has been removed because there is a separate repository for automatic and distributed docking which includes docking with Vina, smina and gnina - https://github.com/ci-lab-cz/easydock.