Computational Methods

We are developing new theoretical methods to improve the efficiency of molecular simulations. On the one hand, we are using large-scale simulations to evaluate the statistical convergence of simulations in order to identify sampling bottlenecks and devise appropriate strategies to overcome these limitations. On the other hand, we are developing new generalized-ensemble algorithms that take advantage of large-scale computational resources. These approaches make replica exchange simulations, the current state-of-the-art for large-scale biomolecular simulations, more practical by removing the need for synchronous simulations on dedicated supercomputers and clusters. The algorithms are suitable for shared and heterogeneous computing platforms such as distributed networks.

These advances are helping us to expand the scope of biomolecular simulations. Applications of these methods are being exploited in our laboratory for free energy calculations and large-scale simulation studies of membrane permeation, protein-ligand binding, peptide-lipid interactions, and biomolecular self-assembly.

Statistical Convergence of Equilibrium Properties in Simulations of Molecular Solutes Embedded in Lipid Bilayers. Neale, C.; Bennett, W. F. D.; Tieleman, D. P.; and Pomès, R. Journal of Chemical Theory and Computation, 2011.

Simulated Tempering Distributed Replica Sampling , Virtual Replica Exchange , and Other Generalized-Ensemble Methods for Conformational Sampling. Rauscher, S., and Neale, C., and Pomès, R. J. Chem. Theory and Comput., 2009.

Equilibrium exchange enhances the convergence rate of umbrella sampling. Neale, C.; Rodinger, T.; and Pomès, R. Chem. Phys. Lett., 2008.