Kripo

Key Representation of Interaction in POckets, see http://dx.doi.org/10.1186/1758-2946-6-S1-O26 for more information.

Command line tool to generate Kripo fingerprints from Protein Data Bank files. The fingerprints can be used to can be used to find proteins which can bind a fragment of a ligand.

The kripo algorithm in a nutshell is:

1. Foreach PDB
2. Foreach ligand
3. Fragment ligand into fragments
4. Determine which residues of protein are in contact with fragment
5. Translate contact residues to pharmacophoric features
6. Convert pharmacophore to three point fingerprint

It uses https://github.com/3D-e-Chem/kripodb to store output in database files.

Usage

The ligands in PDB files have no bond information so we use the ligand expo (http://ligand-expo.rcsb.org) for this. The ligand expo sdf file must be downloaded and indexed using:

wget http://ligand-expo.rcsb.org/dictionaries/all-sdf.sdf.gz
kripo ligands import all-sdf.sdf.gz ligand-expo.db

To generate pharmacophore fingerprints from 2 pdb files use:

echo tests/fixtures/3HEG.pdb > pdb.list
echo tests/fixtures/5IS0.pdb >> pdb.list
kripo generate --ligand-expo ligand-expo.db pdb.list frags.db phars.h5 fingerprints.db

This will generate Kripo fragments/pharmacophores/fingerprints for the PDB files listed in the pdb.list file.

To host the Kripo web service the following steps must be done:

# Fetch Uniprot, Enzyme for each PDB entry
kripodb fragments pdb frags.db
# Calculate similarity scores between fingerprints pairs
kripodb fingerprints similarities --fragmentsdbfn frags.db fingerprints.db fingerprints.db similarities.h5
# Freeze similarity pairs into a matrix which can be quickly queried for most similar
kripodb similarities freeze similarities.h5  similarities.frozen.h5
# Startup the web service
kripodb serve similarities.frozen.h5 frags.db phars.h5

Other commands are described at http://kripodb.readthedocs.io/en/latest/cli.html .

Install

To install run

conda env create -n kripo -f environment.yml
conda activate kripo
python setup.py install

Conda is used to install rdkit, reduce and openbabel.

You will have kripo and kripodb commands available in your PATH.

For development of kripo code use pip install -r requirements_dev.txt instead of python setup.py install.

Reference

KRIPO – a structure-based pharmacophores approach explains polypharmacological effects; Tina Ritschel, Tom JJ Schirris, and Frans GM Russel; J Cheminform. 2014; 6(Suppl 1): O26; Published online 2014 Mar 11; http://dx.doi.org/10.1186/1758-2946-6-S1-O26

For protonation of protein Reduce http://kinemage.biochem.duke.edu/software/reduce.php is used: Word, et. al. (1999) Asparagine and Glutamine: Using Hydrogen Atom Contacts in the Choice of Side-chain Amide Orientation, J. Mol. Biol. 285, 1733-1747.

For protonation of ligand OpenBabel https://doi.org/10.1186/1758-2946-3-33 is used.

Legal

Kripo is copyright 2018 Vrije Unversiteit Amsterdam and Netherlands eScience Center. It is free, open source software. You can redistribute it and/or modify it under the terms of either:

• The GNU General Public License as published by the Free Software Foundation, either version 2 or, (at your option), any later version, or
• The Apache License 2.0