Towards Dynamic Pharmacophore Models by Coarse Grain Molecular Dynamics

Abstract

Pharmacophore models play a key role in computer aided drug discovery e.g. in virtual screening of chemical databases, de novo drug design, and lead optimization. Structure-based methods for developing pharmacophore models are of particular importance, and there have been a number of studies combining such methods with the use of molecular dynamics (MD) simulations to model protein flexibility. At the same time, ongoing developments in multi-scale simulations, in which atomistic and CG MD simulations are combined in a sequential fashion, have been successfully used to explore the interactions of proteins with membranes and their lipids [1,2]. Here we describe the use of CG-MD simulations to explore the interactions of CG-particle probes with proteins, combined with cavity detection methods, in order to identify potential protein binding sites. Using cyclin dependent kinase 2 (CDK2) as a test case, we demonstrate the potential of this approach for identification of ligand binding sites, and as part of an overall process of structure-based development of pharmacophore models.