Accurately predicting how fluctuating protein environments recognize molecular ligands have been a long-pursued challenge in computational chemistry and biophysics. The objective of this project is to develop and employ practical all-atom molecular dynamics simulation methods to study how protein environment changes govern protein-ligand recognition. This project is particularly encouraged by our novel sampling method developments, represented by the orthogonal space sampling scheme, which can uniquely enable synchronous acceleration of the motion of a focused (chemical) change and its coupled environmental responses, for instance conformational transitions and wetting/dewetting processes. This study includes three specific goals: (1) to understand how protein environment changes couple with protein-ligand binding and to realize accurate ranking of ligand relative binding affinities;(2) to enable simultaneous prediction of protein-ligand absolute binding affinities and binding complex structures;and (3) to elucidate how fluctuating enzymes recognize tight transition state analogue binders versus their natural substrates.
Leap-frog efficiency improvements for free energy perturbation calculations will be realized to enable quantitative understanding of how protein environment changes govern protein-ligand recognition. These methods will be utilized to elucidate essential protein-ligand interactions. In addition to specific biophysical understanding, a powerful toolkit for quantitative protein-ligand binding predictions will be produced to be feasibly and widely employed in biochemical studies and drug discovery processes.