Understanding protein interactions and their function needs better structural knowledge at atomic level. However, given the limitations of experimental methods such as NMR and X-ray crystallography, a variety of computational tools have been reported to predict a complex structure by docking. Currently, a major bottle-neck is the treatment of flexibility (cases with large conformational changes upon binding give poor predictive success rates). To this aim, we show here the generation of conformational ensembles for the unbound docking partners using different molecular dynamics techniques, which are used as input structures for our docking program pyDock. The results on a benchmark of 124 protein-protein complexes, indicate that this approach is able to increase pyDock performance. In virtually all cases the docking success rates improved with the best-energy conformer, especially in cases with unbound/bound RMSD between 1.0 and 1.5 �. We discuss the application to complexes of biochemical and therapeutic interest.

Unbound conformational ensembles improve protein-protein docking

243rd National Meeting (2012)

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