Book of Abstracts: Albany 2011

category image Albany 2011
Conversation 17
June 14-18 2011
©Adenine Press (2010)

Efficient Explicit Solvent Simulations of Molecular Association Kinetics

Atomically detailed views of molecular association events are of great interest to a variety of research areas in biology and chemistry. A natural approach to providing such views is to use molecular dynamics (MD) simulations in explicit solvent which are quite routine for tens of nanoseconds in certain systems (1-4). However, it has been computationally prohibitive to perform MD simulations for a sufficiently long time (by “brute force”) to capture more complicated association events, e.g. protein-protein associations, that require microseconds or beyond (5). Fortunately, the long timescales required are not necessarily because the actual events take a long time; instead the events may be fast but infrequent, separated by long waiting times. Path sampling approaches aim to capture rare events by minimizing the simulation of long waiting times between events [as reviewed by Zwier and Chong (6)]. In this work, we have combined the “weighted ensemble” path sampling approach (7) with explicit solvent MD simulations. This approach allows us to obtain accurate kinetics as well as ensembles of molecular association pathways. We have determined the efficiency of this approach relative to brute force simulations in sampling the molecular association events for a range of well-studied systems: methane/methane, Na+/Cl-, methane/benzene, and K+/18-crown-6 ether (pictured below from left to right). Relative to brute force simulation, we obtain efficiency gains of at least 1,100-fold for the most challenging system, K+/18-crown-6 ether, in terms of sampling the distribution of molecular association pathways. Our results indicate that weighted ensemble sampling is likely to allow for even greater efficiencies for more complex systems with higher barriers to molecular association. Applications of weighted ensemble sampling to explicit solvent MD simulations of protein binding events will also be discussed.

Matthew C. Zwier
Joseph W. Kaus
and Lillian T. Chong

Department of Chemistry
University of Pittsburgh
Pittsburgh, PA 15260

Ph: (412) 624-6026
Fx: (412) 624-8301

This research has been supported by NSF CAREER Award MCB-0845216.


  1. P. Sklenovský and M. Otyepka, J Biomol Struct Dyn 27, 521-540 (2010).
  2. M. J. Aman, H. Karauzum, M.G. Bowden and T.L. Nguyen, J Biomol Struct Dyn 28, 1-12 (2010).
  3. C. Koshy, M. Parthiban and R. Sowdhamini. J Biomol Struct Dyn 28, 71-83 (2010).
  4. Y. Tao, Z. H. Rao and S. Q. Liu. J Biomol Struct Dyn 28, 143-158 (2010).
  5. K. A. Henzler-Wildman, D. Kern. Nature 450, 964-972 (2007).
  6. M. C. Zwier, L. T. Chong. Curr. Opin. Pharmacol. 10, 745-752 (2010).
  7. G. A. Huber, S. Kim. Biophys J 70, 97-110 (1996).