Mutants in PCM

This repository contains the open-source package mutants_in_pcm for the analysis of the effect of mutant data in bioactivity modelling, particularly in proteochemometric modelling (PCM). The package is written in Python 3.7 and facilitates the analysis of variants reported in the ChEMBL database and the Papyrus modelling dataset. The full analysis is described in the paper "A comprehensive analysis of the mutant landscape of bioactivity databases and its effect on modeling" by M. Gorostiola González, O. Béquignon et al.

The repository contains an analysis folder with the main script and Jupyter notebooks used for the analysis in the paper. These notebooks can be directly used to quickly and easily extend the analysis to any protein of interest to the user.

Installation

The package can be installed using pip:

pip install git+https://github.com/CDDLeiden/mutants-in-pcm.git@master

Getting started

The package contains three categories of modules:

1. Annotation: modules for the download and annotation of variants in the ChEMBL database and the Papyrus modelling dataset. Includes preprocessing.py, annotation.py, and annotation_check.py.
2. Mutant analysis: modules for the analysis of the effect of variants in bioactivity at different levels. Includes mutant_analysis_family.py, mutant_analysis_accession.py, mutant_analysis_organism.py, mutant_analysis_type.py, mutant_analysis_protein.py, mutant_analysis_compounds.py, mutant_analysis_common_subsets.py, mutant_analysis_distributions.py, and mutant_analysis_clustermaps.py.
3. Modelling: modules for the modelling of bioactivity data, including the modelling of variants, and analysis of the effect of variants. Includes descriptors.py and modelling.py.

Data availability