All-atom computations of long-range interactions with nonreversible Markov chains

All-atom computations of long-range electrostatic interactions with reversible Markov-chain Monte Carlo (MCMC) require a number of computations that increase as N^{3 / 2} with the number of particles N in the system. We recently designed a nonreversible MCMC algorithm for all-atom electrostatic interactions whose number of required computations instead increases as N log(N) for systems composed of charge-neutral molecules, such as liquid water. We're currently developing our method to apply to more complex charged systems such as biological proteins and long-chain charged polymers in ionic batteries.

This is a collaborative project with Werner Krauth, and Liang Qin at Ecole normale supérieure and Tony Maggs at ESPCI ParisTech.

Long-range electrical interactions are key to understanding a broad range of physical phenomena from protein folding in biological cells to ionic fluids in battery technology. We recently designed a new Markov-chain Monte Carlo algorithm for simulating electrical interactions, which we expect to outperform other modern methods when applied to these electrically charged systems.