The accurate prediction of the binding affinity changes of drugs is a major goal in drug discovery and personalised medicine. Our approach, thermodynamic integration with enhanced sampling (TIES), utilises the concept of ensemble molecular dynamics (MD) simulation to yield rapid, accurate, precise and reproducible binding affinities. TIES is designed to compare the binding free energies of congeneric ligands, and perform wells for a wide range of target proteins and ligands. We have recently introduced variants of TIES, which incorporate enhanced sampling techniques such as REST2 (replica exchange with solute tempering) and the free energy estimator MBAR (multi-state Bennett acceptance ratio). We have also extended the methodology to for the computation of binding affinity changes caused by protein mutations when bound to a ligand, a variant which we call it TIES-PM. The reliability of computer simulations is a widespread concern. In MD, one-off simulations are not reproducible due to the chaotic nature of the trajectories. We apply a systematic protocol for uncertainty quantification (UQ) to TIES and its variants, applicable to both absolute and relative free energy predictions. A reliable measure of error estimation is provided by ensemble approaches such as TIES and its variants, which apply irrespective of the free energy method. The need to use ensemble methods is fundamental and holds regardless of the duration of time of the molecular dynamics simulations performed, which we also apply to Schrödinger’s FEP+ in order to make meaningful comparisons. After describing how rigorous UQ is intrinsic to TIES in all its flavours, we report a study of TIES-PM applied to fibroblast growth factor receptor 3 (FGFR3), a protein that plays a critical role in many physiological processes, inter alia being a recognized therapeutic target in cancer treatment. We find that TIES-PM with REST2 successfully captures a large conformational change and generates correct free energy differences caused by the gatekeeper mutation, a common residue variation occurring inside the ligand binding pocket. Simulations performed without REST2, however, fail to overcome energy barriers between conformations and are then senstive to the choice of initial structures.

Ensemble-based molecular dynamics: Uncertainty quantification and enhanced sampling techniques in free-energy calculations

2019 Spring National Meeting

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