Modern alchemical binding free energy calculations involve many choices of simulation parameters. We have focused on methods based on temperature and Hamiltonian Replica Exchange which, among others, requires choosing the form of the alchemical hybrid potential and the soft-core potential, as well as the number of simulation replicas and their distribution along the alchemical progress parameter lambda. Other choices involve setting suitable restraints to focus sampling in specific regions of conformational space, or, alternatively, judiciously broaden sampling to increase the likelihood of finding a favorable binding pose. We will illustrate the application of an analytic model of alchemical molecular binding we recently developed to the problem of finding optimal simulation parameters for binding free energy calculations. We will show how the model can be used to optimize the lambda schedule and help in the design of efficient soft-core energy functions. We will focus in particular on strategies to remove sampling bottlenecks along the alchemical path caused by binding-induced conformational changes and order-disorder pseudo phase transitions.

Optimization of conformational sampling protocols using an analytic model of alchemical molecular binding

2019 Spring National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.