Characterizing protein-protein binding: from predicting protein binding sites with graph theory to calculating hot spot amino acids
Nejc Carl Nejc Carl
Laboratory for Molecular Modeling
Chemical Institute
Abstract:
To better understand the role of amino acids in protein-protein binding we have expanded on an existing quick approach for protein-protein binding site detection, that is based in graph theory. With this approach we are able to find local structural similarities to a protein of interest in tens of thousands of protein structures. We define local structural similarities as patches of surface that have a similar arrangement of functional groups. In order to determine the role of a particular amino-acid in protein-protein binding, we have rigorously calculated their energy of binding with thermodynamic integration. Thus we have combined a quick approximate method with a computationally intensive rigorous method. Here we present a proof of concept results, on a case of protein 1aze.