The method of predicting protein binding sites using small organic molecules as probes has been utilized in a large number of computational experiments such as protein function prediction, determining protein relationships, and drug design. A variety of such algorithms are available, however, most of them focus on finding sites specifically for organic small molecules. We have utilized a Fast Fourier Transform based protein-protein docking approach with accurate molecular mechanics, like potentials, to predict larger protein binding sites. This method utilizes approximate grid based continuous electrostatics solvation model, and has proven successful in test cases. The algorithm uses macromolecules as both target and probe, and measures the interacting energy of the two macromolecules. The lowest energy poses are retained and through several iterations of clustering, return regions of the target protein surface where the binding energy is most frequently at its lowest. These regions frequently correlate with protein binding sites. This information can then be used to refine docking results or determine additional information on the protein binding pocket.

Predicting macromolecular binding sites on protein binding partners

248th National Meeting (2014)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.