Incorporating experimental restraints into parallel ensemble simulation to refine the conformations of flexible proteins

Peter Kasson, Ohio Supercomputer Center

Usage Details

Flexible molecular recognition is a common paradigm in immunity and infection; many pathogens have proteins that are structurally flexible and/or highly tolerant of mutations, but still effectively bind to human cells. Determining the structural basis of this recognition experimentally is challenging because of the multitude of structures involved. We have developed a computational methodology that provides a way to select the best experiments to measure these structures and subsequently combine them in an integrated model. The conformations of many important flexible proteins are still not well understood; this methodology will allow researchers to characterize systems that are currently too difficult to understand with existing refinement techniques. A systematic approach to refining these flexible receptor-ligand complexes would 1) help elucidate the fundamental physical principles of receptor-ligand binding and 2) promote better drug design for infectious disease.