This web-site was created to share some developments in chemoinformatics tools which can be useful for chemoinformaticians as well as for medicinal chemists and other specialists. It contains tools and approaches which were developed in collaboration with my colleagues or by myself.

Research interests:

Applied chemoinformatics:
  • medicinal chemistry projects and development of novel therapeutical agents (antivirals, anti-aggregants, etc) by means of ligand- and structure-based approaches (molecular docking, pharmacophores, etc)
  • QSAR modeling of biological and physico-chemical properties of single compounds and their mixtures and QSAR modeling of chemical reactions
Theoretical studies:
  • development of software tools for QSAR modeling and interpretation

Advances and developments:

  • Simplex representation of molecular structure (SiRMS) - very flexible representation of structures of chemical compounds
  • Representation of mixtures of chemical compounds (based on SiRMS) - for QSAR modeling and prediction of properties of mixtures of chemicals
  • Quasi-mixture representation of individual chemical compounds (based on SiRMS) - improves performance of QSAR models of different physico-chemical end-points
  • Mixture representation of chemical reactions (based on SiRMS) - encodes reactions without explicit mapping of the reaction centre
  • Structural interpretation of QSAR models - universal approach for structural interpretation of QSAR models regardless machine learning method and descriptors used
  • Physico-chemical interpretation of QSAR models - approach for interpretation of QSAR models in terms physico-chemical properties of structural motifs and deos not depend on machine learning method used
  • SPCI - knowledge-mining tool to retrieve SAR from chemical datasets based on structural and physico-chemical interpretation of QSAR models
  • SiRMS - simple tool for generation of 2D SiRMS descriptors for single compounds, mixtures, quasi-mixtures and chemical reactions
© Pavel Polishchuk 2010-2016