Construction of Coarse-Grained Models by Reproducing Equilibrium Probability Density Function

The present work proposes a novel methodology for constructing coarse-grained(CG) models, which aims at minimizing the difference between CG model and the corresponding original system. The difference is defined as a functional of their equilibrium conformational probability densities, then is estimated from equilibrium averages of many independent physical quantities denoted as basis functions. An orthonormalization strategy is adopted to get the independent basis functions from sufficiently preselected interesting physical quantities of the system. Thus the current method is named as probability density matching coarse-graining(PMCG) scheme, which effectively takes into account the overall characteristics of the original systems to construct CG model, and it is a natural improvement of the usual CG scheme wherein some physical quantities are intuitively chosen without considering their correlations. We verify the general PMCG framework in constructing a one-site CG water model from TIP3 P model. Both structure of liquids and pressure of the TIP3 P water system are found to be well reproduced at the same time in the constructed CG model.