Novel Computational Methods for Modeling Cytochrome P450 Mediated Drug Metabolism

Purdue University, West Lafayette IN

Investigators

Linked publications & trials

Abstract

DESCRIPTION (provided by applicant): The metabolism of drugs and other xenobiotics by cytochrome P450 enzymes (CYP) is an essential detoxification and drug clearance mechanism in humans. Despite the publication of several X-ray structures since 2000, reliable computational prediction of drug metabolism remains a huge challenge. In this grant application, we propose an integrative high-throughput approach that combines electronic properties of a ligand with structural properties of the protein. Our main aim is the development and application of innovative structure-based design techniques that address serious shortcomings of current approaches. In particular, we have extended our novel docking concept to incorporate all observed forms of protein flexibility relevant for ligand-CYP interactions, entropic contributions influencing the prediction of binding poses, and we will add on- the-fly solvation to ligand-CYP complexes. To improve the docking quality we will optimize the parameters of a scoring function tailor-made for each CYP enzyme studied. In combination with an initial focus on efficient calculation of hydrogen-abstraction energies we will predict regioselective metabolism of drugs or drug candidates binding to CYPs. Based on the resulting docking poses, multidimensional QSAR simulations will be performed for accurate quantification of binding affinity as a measure CYP-inhibition and better ranking of binding modes. The new computational methods will be applied to two CYP enzymes (CYP2C9, 3A4) important in drug metabolism. The generated computational models will be stored in a database and made publicly available. Other researchers will be invited to screen compounds against our CYP database via a secure Web protocol to predict drug metabolism and inhibition. The submission of data by other researchers will provide valuable feedback on the performance and applicability of the models.