It is well accepted that water plays a crucial role in mediating the interactions between ligands and their associated macromolecular receptor. In some cases these water molecules may be buried in protein cavities and unable to exchange readily with bulk solvent. In the context of protein-ligand binding free energy calculations, this creates particular difficulty when the number or binding locations of these trapped waters depends on the ligand. If the water molecules are unable to equilibrate over the course of the molecular simulation, then clearly the free energies calculated will likely be in error.<br/><br/>We have recently implemented and extended the grand canonical Monte Carlo (GCMC) simulation approach to protein-ligand systems. Using a single set of simulations, our methods can locate water molecules, estimate their binding affinities, capture the cooperativity of the water network, and evaluate the hydration free energy of entire protein binding sites. In this presentation we report the further extension of GCMC by combining it with alchemical free energy calculations. In this way the number and location of water molecules in a protein-ligand complex are able to dynamically adjust to the ligand through trial insertion and deletion moves. The results of this new methodology will be compared with more conventional, and expensive, thermodynamic cycle approaches where double decoupling is used to capture changes in free energy arising from the displacement of bound water molecules. The implications of this methodology in the context of automated free energy calculations will be discussed

Incorporating changes in protein-ligand hydration in free energy calculations

251st National Meeting (2016)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.