Protein family- and proteome-wide characterization of drug binding interactions for polypharmacology
Protein binding sites are critical for protein function and are much more conserved that sequences or domain structures. We are working on a number of projects to systematically analyze and compare drug target binding sites – among protein families, across the druggable proteome or across species. Using a number of different computational approaches we systematically model protein families or drug targets covering the entire proteome and derive similarity relationships based on physicochemical properties of binding sites and protein-ligand interaction; we term this ‘phylostructeomics’ analyses. The goal of this work is to facilitate proteome-wide structure-based design and to identify proteins that can potentially be targeted together or to prioritize possible ‘off-target’ effects that should be experimentally tested. The kinase analog of the human proteome (Kinome) is one or the largest protein families. Kinases are involved in many signal transduction pathways and therefore represent one of the major target classes of pharmaceutical drug discovery programs. Protein phosphatases – catalyzing the complementary reaction of kinases – are similarly involved in many physiological processes. The coordinated action of different kinases and phophatases is a powerful and robust mechanism to control protein phosphorylation and regulate these cellular processes. In one of our projects we are developing computational structural biology-based platforms focused on the kinome and tyrosine phosphatome respectively. Because of the broad implications of these signaling proteins, the kinome and tyrosine phosphatome platforms,and ATP-binding proteins are widely applicable to biological and pharmacological problems. In another project we are analyzing human vs. bacterial ATP-binding proteins to develop poly-pharmacological approaches to target parasites without affecting humans.