Constant-pH Hybrid Nonequilibrium Molecular Dynamics鈥揗onte Carlo Simulation Method

详细信息    查看全文

• 作者：Yunjie Chen ; Beno卯t Roux
• 刊名：Journal of Chemical Theory and Computation
• 出版年：2015
• 出版时间：August 11, 2015
• 年：2015
• 卷：11
• 期：8
• 页码：3919-3931
• 全文大小：476K
• ISSN：1549-9626

文摘

A computational method is developed to carry out explicit solvent simulations of complex molecular systems under conditions of constant pH. In constant-pH simulations, preidentified ionizable sites are allowed to spontaneously protonate and deprotonate as a function of time in response to the environment and the imposed pH. The method, based on a hybrid scheme originally proposed by H. A. Stern (J. Chem. Phys. 2007, 126, 164112), consists of carrying out short nonequilibrium molecular dynamics (neMD) switching trajectories to generate physically plausible configurations with changed protonation states that are subsequently accepted or rejected according to a Metropolis Monte Carlo (MC) criterion. To ensure microscopic detailed balance arising from such nonequilibrium switches, the atomic momenta are altered according to the symmetric two-ends momentum reversal prescription. To achieve higher efficiency, the original neMD鈥揗C scheme is separated into two steps, reducing the need for generating a large number of unproductive and costly nonequilibrium trajectories. In the first step, the protonation state of a site is randomly attributed via a Metropolis MC process on the basis of an intrinsic pK_a; an attempted nonequilibrium switch is generated only if this change in protonation state is accepted. This hybrid two-step inherent pK_a neMD鈥揗C simulation method is tested with single amino acids in solution (Asp, Glu, and His) and then applied to turkey ovomucoid third domain and hen egg-white lysozyme. Because of the simple linear increase in the computational cost relative to the number of titratable sites, the present method is naturally able to treat extremely large systems.