/CG4-Predictor

Predictor of ChemGauss4 docking score from molecular features

Primary LanguageJupyter Notebook

ChemGauss4 docking score predictor based on molecular features

Based on results of a virtual screen against an integrin homology model

Takes input in the form:

[Polar surface area (ang^2), Hydrogen bond acceptor groups, hydrogen bond donor groups, largest ring size, molecular weight (gmol^-1), number of ring systems, number of rotatable bonds]

Limitations / ToDO:

  • Use a xray structure to acquire docking data
  • Save more than 200 ligands per ZINC15 tranche for representative sample of scores
  • Optimize network geometry