United-Atom Discrete Molecular Dynamics of Proteins Using Physics-Based Potentials

Abstract

We present a method for the efficient simulation of the equilibrium dynamics of proteins based on the well established discrete molecular dynamics algorithm, which avoids integration of Newton equations of motion at short time steps, allowing then the derivation of very large trajectories for proteins with a reduced computational cost. In the presented implementation we used an all heavy-atoms description of proteins, with simple potentials describing the conformational region around the experimental structure based on local physical interactions (covalent structure, hydrogen bonds, hydrophobic contacts, solvation, steric hindrance, and bulk dispersion interactions). The method shows a good ability to describe the flexibility of 33 diverse proteins in water as determined by atomistic molecular dynamics simulation and can be useful for massive simulation of proteins in crowded environments or for refinement of protein structure in large complexes.