Multigranular Molecular Dynamics Simulation

This Phase I SBIR project will demonstrate the feasibility of an innovative computational tool for the development of new materials. It has always been the goal of computational chemistry to predict the formation, behavior and properties of a molecule before it is synthesized in the laboratory. While considerable progress has been made, severe computational limitations still inhibit a global solution approach. At one extreme, the quantum codes are needed to predict electronic and optical properties; at the other, molecular mechanics tools are necessary to determine thermal and mechanical characteristics. As long as these extremes are largely uncoupled, the codes are generally useful. Where localized quantum effects drive major macromolecular conformation changes, as in photon absorption by bacteriorhodopsin, conventional codes become inadequate. Recent algorithmic developments by Moldyn, based on aggregation approaches, will however, respond to these needs. Multigranularity is already inherent in aggregation, and the primary task is to include the quantum capability in that formalism. In Phase I, the method will be demonstrated using existing codes - the quantum and atomistic capabilities in CHARMM combined with the aggregation approaches within MBO(N)D. Bacteriorhodopsin will be the test example. In Phase II, the prototype of an end-to-end multigranular code will be assembled.