Long-time methods for molecular dynamics simulations: Markov State Models and Milestoning.

Area of Science:

• Computational Chemistry and Biophysics
• Biomolecular Simulations
• Statistical Mechanics

Background:

• Molecular dynamics (MD) simulations are crucial for studying biomolecules but face limitations in simulating complex systems over long timescales.
• The high-dimensional free energy landscape of biomolecular systems necessitates accurate sampling for identifying kinetic/thermodynamic states and calculating transition rates.
• Computational hardware advancements alone cannot overcome the challenges of simulating large biomolecular systems for extended durations.

Purpose of the Study:

• To review and present applications of two advanced methods, Markov State Models (MSMs) and Milestoning, for long-time molecular dynamics simulations.
• To highlight how these methods enhance the study of complex biochemical systems by improving sampling efficiency and accuracy.
• To demonstrate the potential of these techniques in opening new avenues for biomolecular mechanism investigations.

Main Methods:

• Markov State Models (MSMs): Identify long-lived configuration states and use simulations to map energy landscapes, extracting thermodynamic and kinetic data.
• Milestoning: Enables accurate studies of pathways connecting specific end-states (e.g., reactants and products).
• Both methods leverage sets of short simulations for enhanced sampling efficiency compared to long, continuous MD trajectories.

Main Results:

• MSMs facilitate systematic and automated unveiling of biomolecular mechanisms and accurate estimation of free energy barriers.
• Milestoning provides systematic, accurate, and automated studies of pathway ensembles connecting distinct states.
• Applications include conformational dynamics, peptide binding (amyloid-forming, cell-penetrating), and kinase dynamics (DFG-flip).

Conclusions:

• Markov State Models and Milestoning significantly advance the capability for long-timescale molecular dynamics simulations of biomolecular systems.
• These methods offer systematic, accurate, and automated approaches to investigate complex reaction pathways and mechanisms.
• The increasing adoption of MSMs and Milestoning promises new research opportunities beyond the limitations of current computational hardware.