Statistical mechanics theory of molecular recognition and pharmaceutical design

Abstract

Molecular recognition (MR) is an essential elementary process allowing biomolecules to perform their function. MR can be defined as a molecular process in which one or several guest molecules are bound with a high probability at a particular site such as a cleft or a cavity, of a host molecule in a particular orientation. It is a thermodynamic process which is characterised by the difference of the free energies between two states of a host–guest system, bound and unbound. The process features an extremely heterogeneous atomic-environment around binding sites, which has turned away challenges by the conventional statistical mechanics of liquids, e.g. a mean field theory. We have been developing a new theory for MR in biomolecular systems, based on the statistical mechanics of liquids, or the 3D-reference interaction site model (RISM)/RISM theory. The theory has demonstrated its amazing capability of predicting the process from the first principle. In this article, we review our recent works on MR concerning protein and deoxyribonucleic acid. Some applications of the method to drug design are also presented.