New computational tools at the molecular scale for protein-ligand binding in drug discovery

Abstract:
Proteins interact with other molecules through their protein binding sites. Each binding site usually binds one or more ligands. Detection and characterization of binding sites gives insight in protein functionality and is hence essential for drug design.
We have developed new methodological solutions for the prediction and study of protein binding sites using graph theoretical approaches and molecular dynamics simulations. In particular, we have developed computational tools – ProBiS tools – which enable drug discovery based on protein structures (ProBiS, ProBiSCHARMMing, GenProBiS and ProBiS H2O web servers). The GenProBiS web server implements a novel approach to the discovery of sequence variants that have potentially deleterious effect on protein function and ligand binding through gain or loss of the binding site. A novel ProBiS H2O approach uses existing experimental structural data to identify conserved water sites on the interfaces of protein complexes, for example protein–small molecule interfaces, and elsewhere on the protein structures. Our tools enable joining several otherwise disconnected areas of research, for example genome sequence studies, protein structures, and MD simulations.