PyDock is an energy-based protein-protein docking protocol with excellent results in several CAPRI rounds. Predictions become more challenging when dealing with conformational changes upon binding. Explicit protein flexibility has a high computational cost. An inexpensive alternative is the use of low resolution data, less sensitive to small structural changes. Here we present different approaches to protein docking using low resolution data and their integration into our all-atom docking protocol. We apply a small-network approach to the representation of protein structures, showing that topological features of the resulting graphs include valuable information to guide protein docking. We also derive intermolecular residue contact preferences from a manually curated database of protein heterocomplexes. We combine this potential with a coarse-grained desolvation term to define a new scoring function that outperforms our all-atom scoring function. Finally, we show how SAXS experimental data can be combined with protein docking, overcoming their individual limitations.

Integration of low-resolution data into all-atom protein docking

243rd National Meeting (2012)

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