Thursday, November 30, 20172:00 PM - 3:00 PMCNLS Conference Room (TA-3, Bldg 1690)|
Accelerating molecular simulations of lipid bilayers
Sean Marks T-CNLS
The local structure of cell membranes can determine the fate of embedded biomolecules that govern cell signaling. Such membranes are mixtures of various lipids--macromolecules with hydrophilic head groups and hydrophobic tails--that assemble into bilayers with complex local structure. It is this local structure that influences the behavior of biomolecules that transmit intercellular signals, including those associated with many cancers. For example, the Ras family of proto-oncogene proteins contains some members which partition into locally ordered regions, while other members prefer locally disordered regions. While numerous experimental studies of lipid bilayers have been performed, much remains unclear about the precise nanoscopic structure which determines the partitioning behavior of embedded molecules.
Molecular dynamics (MD) simulations can provide valuable insight regarding lipid mixtures at length scales currently unreachable by experiments. Yet advancement on this front has been greatly hindered by the exceptionally slow time scale associated with lipid diffusion. To address this issue, we have applied the enhanced sampling method known as Replica Exchange with Solute Tempering (REST) to the study a ternary lipid mixture. We observe that REST speeds up equilibration by an order of magnitude versus standard MD simulations. Accordingly, we believe that REST has the potential to dramatically enhance the productivity of molecular simulations of lipid bilayers, including those with membrane proteins.
Host: Angel Garcia