Determining energy barriers and selectivities of a multi-pathway system with infrequent metadynamics

Estimating the transition rates and selectivity of multi-pathway systems with molecular dynamics simulations is expensive and often requires arduous sampling of many individual pathways. Developing a way to efficiently sample and characterize multi-pathway systems creates an opportunity to apply these tools to study systems that, previously, would have had a prohibitive computational cost. We present an approach that places quartic boundaries at the saddle points to isolate individual pathways without changing their observed rates, reducing the required number of events sampled and estimated rate uncertainty. In addition to rates, the selectivity between pathways is also accurately predicted as well. To further reduce the computational cost of the analysis, we have paired this approach with the infrequent metadynamics method. The method is demonstrated on model systems and stiffened alanine dipeptide. Furthermore, we present an appropriate method for recovering the energy barriers of specific transition paths by taking the slope of an Arrhenius plot generated from the infrequent metadynamics results at various temperatures. We also compare this method against another previously published literature to demonstrate its superior performance. In the future, these methods can be used in a variety of contexts where competing escape pathways with different barriers are relevant.