One of the computational grand challenge problems is to develop methodology capable of sampling conformational equilibria in systems with rough energy landscapes. If met, many important problems, most notably protein structure prediction, could be significantly impacted. In this talk, I will present a new approach [1,2] in which molecular dynamics is combined with a novel variable transformation designed to warp configuration space in such a way that barriers are reduced and attractive basins stretched. The new method rigorously preserves equilibrium properties while leading to very large enhancements in sampling efficiency. The performance of the method is demonstrated on long polymer chains and simple protein models and is shown to significantly outperform replica-exchange Monte Carlo with only one trajectory [2]. Finally, a new molecular dynamics approach for generateing multi-dimensional free-energy surfaces that employs adiabatic dynamics [3,4] combined with multiple time-step integration to drive a set of extended phase-space variables as a means of enhancing the sampling of a subspace of collective variables will be presented.

References:

1. Z. W. Zhu, M. E. Tuckerman, S. O. Samuelson and

G. J. Martyna, Phys. Rev. Lett. 88:100201 (2002).

2. P. Minary, M. E. Tuckerman and G. J. Martyna,

SIAM J. Sci. Comput. (in press).

3. L. Rosso, P. Minary, Z. W. Zhu and M. E. Tuckerman,

J. Chem. Phys. 116:4389 (2002).

4. J. B. Abrams, L. Rosso and M. E. Tuckerman,

J. Chem. Phys. 125:074115 (2006).