Estimating Full Path Lengths and Kinetics from Partial Path Transition Interface Sampling Simulations.

Area of Science:

• Computational Chemistry
• Biophysics
• Statistical Mechanics

Background:

• Molecular dynamics (MD) simulations face challenges in accurately assessing biological process timescales due to rare and slow events.
• Existing methods struggle to extract time-dependent properties from computationally efficient, but short, simulation paths.

Purpose of the Study:

• To develop a formalism for extracting kinetic information from partial paths generated by the replica exchange partial path transition interface sampling (REPPTIS) algorithm.
• To enable accurate estimation of time-dependent properties like mean first passage times and rate constants from REPPTIS simulations.

Main Methods:

• Introduction of a Markov state model (MSM) framework to analyze overlapping partial paths from REPPTIS.
• Derivation of closed-form formulas for crossing probability, mean first passage times (MFPTs), flux, and rate constants within the MSM framework.
• Validation using simulations of Brownian/Langevin particles and KCl dissociation, and application to the trypsin-benzamidine complex.

Main Results:

• The developed MSM framework successfully estimates full path lengths and kinetic properties from REPPTIS partial paths.
• Validated accuracy by reproducing benchmark kinetics for physical systems and estimating dissociation rates for a biological complex.
• Demonstrated the capability to extract crucial kinetic information from computationally efficient simulations.

Conclusions:

• The MSM framework provides a robust theoretical and practical foundation for REPPTIS simulations.
• Enables the extraction of reliable kinetic information from computationally efficient partial paths, overcoming limitations of previous methods.
• Advances the study of rare and slow events in molecular dynamics simulations of biological processes.