Replica Exchange Monte Carlo GRID: A novel high-resolution refinement algorithm

Abstract

The energy-based refinement of protein structures to atomic-level accuracy remains a major challenge in structural biology. Energy-based refinement is mainly dependent on two components: (1) sufficiently accurate force fields, and (2) efficient conformational space search algorithms. Focusing on the latter, we developed a high-resolution Replica Exchange Monte Carlo-based refinement algorithm called REMC-GRID. This method takes a three-dimensional protein structure as input and, employing an all-atom force field, attempts to improve the energy of the structure by randomly selecting residues and perturbing the backbone dihedral angles. GRID, another refinement algorithm that we had developed previously, similarly improves the energy of protein structures, but is deterministic and perturbs the backbone dihedrals and conformation of all the residues in a sequential fashion. We applied REMC-GRID and GRID to 10 high-resolution (≤ 2.8 Å) crystal structures from the Protein Data Bank and measured the energy improvements obtained and the computation times required to achieve them. REMC-GRID produced better energy improvements than GRID alone and was only moderately more expensive in the use of computational resources. In another set of experiments, we created decoy structures by randomly perturbing the backbone dihedrals, and then tested the ability of the two algorithms to return the structures back to the native conformation. REMC-GRID resulted in better recapitulation of the native conformation than GRID as measured by backbone RMSD.

Additional details