Introduction to QSAR modeling based on RDKit and Python.

This tutorial was created for the Advanced in silico drug design workshop, 30 January - 1 February 2017 (Palacky University, Olomouc, Czech Republic).

Topics covered:

  • Reading SDF file
  • Descriptor calculation and scaling
  • Creation of training and test sets and cross-validation folds
  • Model building and tuning on examples of RF, SVM anf GBM models
  • Prediction
  • Applicability domain
  • Consensus modeling
  • Model interpretation: variable importance


Html-version of Jupyter notebook

Jupyter notebook

Archive with the notebook and all input/output files

© Pavel Polishchuk 2010-2019