527 Applications of Enhanced Sampling Methods in Molecular Dynamics Simulations of Complex Biological Systems

Albert Pan, D. E. Shaw Research
The function of complex biological systems is intimately tied to transitions among conformational states, and characterizing these states and their populations is a problem of fundamental importance. In principle, molecular dynamics (MD) simulations can provide this information by sampling different protein conformations with atomic detail. In practice, however, such simulations must sequentially resolve motions occurring at the femtosecond timescale, whereas interesting events such as protein folding, major conformational transitions, and ligand binding typically occur on timescales many orders of magnitude longer. Long and computationally expensive simulations are thus required, and it is sometimes infeasible to adequately sample all significantly populated conformational states during the course of a conventional MD simulation. This limitation has motivated the development of enhanced sampling simulation methods — purely algorithmic changes to conventional MD that aim to accelerate the sampling of conformational states. This mini-symposium is intended to explore the current state of the art in applications of enhanced sampling methods in MD simulations of complex biological systems, as well as other related themes, such as: Which methods work best for sampling protein conformational states? How generally applicable are such methods to different systems? Does the dramatic increase in the power of conventional MD in recent years from massively scalable MD software and from simulations on graphics processing units (GPUs) and special-purpose supercomputers open up opportunities for the design of novel enhanced sampling methods that take advantage of this increased computational power?