(476i) Approaches to Finding Optimal Pathways and Flux Along Them in Multidimensional Free-Energy Hypersurfaces

Barden, D. R., Univ of New Hampshire
Molecular-scale investigations of complex (bio)physical systems necessitate rationalizing reaction mechanisms by obtaining numerical estimates of free-energy differences between all conformational states of interest. To account for many relevant degrees of freedom in flexible molecules, it is often desired to construct free-energy hypersurfaces in a large yet finite number of variables because the knowledge of these hypersurfaces is useful in obtaining information on transition pathways and reactive fluxes along them. Contrary to several path-based simulation methods in which an initial guess of a parameterized path is iteratively optimized to obtain minimal free-energy pathways, in this work we explore initial applications of new simulation methods that allow searches within multidimensional free-energy surfaces by propagating solutions normal to a seed point to obtain the first exit points and hence pathways of minimal resistance. We will present results from the applications of these methods to model-potentials and analytically reconstructed free-energy hypersurfaces.