Algorithms for the simulation of short and long time dynamics of proteins

Trustees Of Boston University, Boston

Investigators

Abstract

John Straub of Boston University is co-supported by a grant from the Theoretical and Computational Chemistry and Molecular Biophysics Programs to continue his efforts at developing computational methods that enable the accurate description of protein dynamics on both long and short time scales. Such techniques must identify and isolate thermodynamically important low lying energy basins and effectively sample those basins and move between them in a way that produces exact thermodynamic averages. A current emphasis is on improving the means for proper accounting of solvation energies that influence protein configurational distributions. Two specific applications are studied. First, with respect to vibrational energy flow, current interest is on exploring the nature of heme cooling as function of protein type and solvent environment. Second, a coarse-grained potential energy function is being improved to enable studies on lysozyme folding dynamics. Methods used for these studies include Monte Carlo simulations, path-integral methods, molecular dynamics and quantum-chemical methods The ability to computationally understand biological processes requires methods that can accurately describe processes that occur on vibrational time scales and protein folding which requires longer time scales. Predictive capabilities on the vibrational time scale allows one to predict relaxation kinetics and directly test such predictions by comparing aspects of theoretically deduced vibrational spectra to experimentally determined results. Computational deduction of protein-folding mechanisms and pathways requires accurate and efficient coarse-grained models that are significantly more complex then those that are currently available. The new techniques being developed and advanced by Straub are beginning to be used for simulating chemical and biophysical processes. Further improvements will significantly impact the field of computational biology.