Modelling variable protonation states in protein-ligand systems

Molecular simulations are now being widely used to understand biological structure and function. An area of particular interest is in calculating protein-ligand binding free energies; these approaches are being increasingly used in the context of pharmaceutical drug discovery. However, they are limited by two fundamental problems - the reliability of the underlying model used to describe the energy and forces in the system, and the ability of the simulation algorithms to sample effectively the configurations of the system.
To address the second of these problems, we have developed our own enhanced sampling approaches, and tested those of others, to improve the sampling of rare conformations of the system. In this context, we are becoming increasingly interested in non-equilibrium approaches, so-called non-equilibrium candidate Monte Carlo (NCMC), where perturbations are made to the system, and the work associated with the relaxation process is used to estimate the associated change in equilibrium free energy. In this project we will use NCMC to model changes in protonation states associated with protein-ligand binding.
We will examine the effect of dynamic protonation states within a protein's active site and how this affects binding of small molecules such as potential drug candidates. The ultimate goal of this project is to develop, implement and test NCMC-based protonation methodology in the context of protein-drug binding free energy calculations, allowing putative drug candidates to be developed with greater speed and accuracy.