Biomolecular transitions: efficient computation of pathways, free energies, and rates

Abstract

We present an efficient method to compute transition rates between states for a two-state system. The method utilizes the equivalence between steady-state flux and mean first passage rate for such systems. More specifically, the procedure divides the configurational space into smaller regions and equilibrates trajectories within each region efficiently. The equilibrated conditional probabilities between each pair of regions lead to transition rates between the two states. We apply the procedure to a non-trivial coarse-grained model of a 70 residue section of the calcium binding protein, calmodulin. The procedure yields a significant increase in efficiency compared to brute-force simulations, and this efficiency increases dramatically with a decrease in temperature.