Title: Simulation of Long Timescale Dynamics of Proteins Using Normal Mode Langevin

Abstract:
The Normal Mode Langevin dynamics integrator (NML) is a novel method that approximates the kinetics or thermodynamics of a biomolecule by a reduced model based on a normal mode decomposition of the dynamical space. The basis set uses the eigenvectors of a mass re-weighted Hessian matrix, calculated with a biomolecular force field. This set is ordered according to the eigenvalues, which has a physical meaning (square-root of the mode frequency). The low frequency eigenvalues correspond to more collective motions, whereas the highest frequency eigenvalues are the limiting factor for the stability of the integrator.

In this method the higher frequency modes are minimized while respecting the subspace of low frequency dynamical modes, which are accurately propagated. The method periodically recomputes the normal modes using a coarse-grained normal mode analysis (CNMA) that captures the essential features of the low frequency modes. CNMA scales as O(N log N) time and O(N) memory, thus retaining the potential efficiency increase of the underlying NML method. NML is capable of accommodating large conformational change of proteins, as shown in simulation of WW domain folding. For this example NML is capable of using time steps of 50 to 100 fs, with real speedups of 17 to 33 over plain Langevin dynamics. Speedups are larger for larger proteins.