181 Discovery of potential inhibitors of BMX non-receptor tyrosine kinase through e-pharmacophore based virtual screening.

Cellular functioning depends on proper coordination between proteins, and much of this communication is mediated by so-called peptide-protein interactions in which a short linear motif is bound by a corresponding peptide-binding domain. These interactions provide major challenges both to experiment as well as modeling due to their often weak, transient character, but our ability to study them is critical for improved understanding, and precise modulation, of cellular regulation. A major focus of our research lies in the development and application of protocols for the accurate modeling of peptide-protein interactions. Rosetta FlexPepDock takes as input a receptor structure and an approximate starting structure of the peptide, and refines this structure to near-native resolution using a Monte-Carlo sampling approach coupled with minimization that starts with an attraction–dominated energy function and gradually increases repulsion to generate well-packed, clash-free models (Raveh et al., 2010, 2011; London et al., 2011a). These models can then be used to rank different peptide sequences for their ability to bind to a given receptor (London et al., 2011b). In my talk I will describe recent new developments and applications of peptide-protein modeling, including initial and advanced steps towards the full ab initio modeling of peptide-protein interactions in which no prior information about the peptide-binding site is available. Most exciting, for peptides with characterized binding motifs, we show that a motif-based search for fragments in solved structures and their subsequent docking allows us to identify an acceptable model of interaction, which can then be further refined using e.g. FlexPepDock. We anticipate that this novel approach and implementation will significantly extend the number of peptide-mediated interactions that can be accurately modeled, and thus characterized and manipulated.