QPHAR: quantitative pharmacophore activity relationship: method and validation

Abstract QSAR methods are widely applied in the drug discovery process, both hit?to?lead and lead optimization phase, as well drug-approval process. Most algorithms limited to using molecules input disregard pharmacophores or pharmacophoric features entirely. However, due high level of abstraction, pharmacophore representations provide some advantageous properties for building quantitative SAR models. The abstract depiction molecular interactions avoids a bias towards overrepresented functional groups small datasets. Furthermore, well?crafted model can generalise underrepresented even missing training set by interaction patterns only. This paper presents novel method construct models demonstrates its applicability robustness on more than 250 diverse fivefold cross-validation these datasets with default settings yielded an average RMSE 0.62, standard deviation 0.18. Additional studies 15–20 samples showed that robust could be obtained. These low requirements dataset sizes render viable go-tomethod medicinal chemists, especially lead-optimisation stage projects.