QSAR approaches to predict human cytochrome P450 inhibition

This thesis focuses on several aspects of QSAR modeling of human cytochrome P450 inhibition and suggests the methodology to increase the quality of CYP inhibition models. It is shown that the addition of newly developed descriptors derived from docking simulations increases the predictive ability of the resulting models. The studies were performed on the OCHEM platform (http://ochem.eu) and all the descriptors, datasets and models are publicly available to the scientific community.