The investigator aims to develop reliable, theoretically coherent computational models for non-covalent binding, focusing on the association of proteins and small, drug-like molecules. Specifically, this will involve (1) Elaboration of the thermodynamic theory of binding; (2) Development and testing of a novel numerical approach to evaluating configuration integrals using the predominant states approximation (3) Testing of models of binding by the ability of an energy function to pick out the correct bound conformation of a ligand from many plausible alternatives. (4) Development and testing of improved models of solvation in binding calculations using continuum Poisson-Boltzmann solvent models. (5) Evaluation of simple models of binding to serve as benchmarks so that a more complex model will only be used if more accurate. (6) Optimization of a new, accurate method for predicting protonation states of titratable groups in proteins, and use in binding calculations.
