Canonical ensemble configurations for explicit water using a nondynamical buildup strategy

PHYS 354

Divesh Bhatt, divesh.bhatt@gmail.com and Daniel M. Zuckerman, dmz@ccbb.pitt.edu. Department of Computational Biology, University of Pittsburgh, 3088 Biomedical Sciences Tower 3, 3501 Fifth Ave, Pittsburgh, PA 15260
Configurations of water in the canonical ensemble are generated using a non-dynamical sampling strategy rooted in well-established polymer growth algorithms. In this strategy, subsequent water molecules are added using a combinatorial approachbased on the preferred orientations with respect to already sampled water configurations.The process is repeated, with reweighting to maintain a canonical distribution, untilequilibrium configurations of the full system are generated. The calculation also yields the absolute free energy. The results are compared with conventional canonical Monte Carlo simulations of pure water. The methodology can, in principle, convert an implicit-solvent ensemble of a flexible solute to the explicit-solvent ensemble -- thus providing a bridge between the two solvent descriptions.