ITR: Accelerated Algorithms for Multiple Time Scale Simulations of Ultraslow Systems: Supercooled Liquids and Biomolecules

Trustees Of Boston University, Boston

Investigators

Abstract

Thomas Keyes, John Straub, and Steven Homer of Boston University are supported by the Chemistry Division and the Division of Materials Research under the Information Technology Research program to develop and apply algorithms for multiple time scale simulations of supercooled liquids and biomolecules. Three different algorithms are being developed for acceleration of transitions between different local minima. All three depend on methods for identifying local minima, transition barriers, and pathways in complicated molecular systems. Biologically interesting processes may be characterized by slow transitions between dissimilar molecular configurations. Dynamical processes on a much faster time scale occur rapidly within each of the configurations and during the slow transition between the two configurations. A simultaneous accurate accounting of both slow and fast processes represents a significant challenge to computational chemists and materials scientists. This work merges techniques for mathematically describing landscapes near a local minimum with information technology methods to allow for modeling of such slow events. The research is strongly coupled to educational activities that are designed to expose young researchers to the multiple time scale problem.