Integrative structural modeling of protein-protein complexes is becoming an emergent approach to interpreting multi-technique biophysical measurements for structure determination. An optimal, integral use of these biophysical data largely depends on the ability of sampling the conformational space of protein-protein interactions. Traditionally, molecular dynamics simulations have been used as a powerful tool for flexible protein docking; however, it remains elusive how to achieve systematic, effective sampling of the vast conformational space. First, in order to make the flexible-docking more computationally tractable, we describe a protein flexible-docking scheme (termed RotPPR) to provide a systematic and exhaustive sampling of relevant translational and rotational degrees of freedom. This is achieved by parallelizing translation-centric, pull-push-release simulations via the use of a rotationally-uniform starting pose generator that was originally developed for rigid-body docking. Second, for analyzing post-simulation data, a novel orientation-specific RMSD (oRMSD) clustering algorithm is introduced and compared with a common RMSD-based method for a proper conformational identification from simulation trajectories. As a result, an energy landscape view of the entire conformational space is enabled by RotPPR sampling and oRMSD clustering. Third, a multi-technique structure modeling is described via the integration of information from small-angle X-ray scattering, hydroxyl radical protein footprinting, and exhaustive RotPPR simulations, termed iSPOT. We show that despite the failure of standalone computations for accurate structure prediction, this hybrid approach of iSPOT modeling is able to accurately narrow down the target structures of more than 15 protein-protein complexes tested in a consistent manner.

Integrated structural modeling of protein-protein complexes assisted by comprehensive, conformational sampling

251st National Meeting (2016)

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