Adaptive Markov state model estimation using short reseeding trajectories

Area of Science:

• Computational chemistry and biophysics
• Molecular dynamics simulations
• Statistical mechanics

Background:

• Advances in molecular dynamics (MD) and Markov State Models (MSM) enable accurate estimation of kinetic rates and pathways for slow biomolecular dynamics.
• Enhanced sampling methods are crucial for efficiently exploring complex conformational landscapes.

Purpose of the Study:

• To investigate the performance of various MSM estimators when applied to reseeding trajectory data.
• To identify practical challenges and propose solutions for adaptive sampling in biomolecular simulations.

Main Methods:

• Application of MSM estimators to reseeding trajectory data.
• Analysis of a 1D free energy landscape.
• Simulation of mini-protein folding for WW domain and NTL9(1-39).

Main Results:

• Reseeding simulations present practical challenges for accurate kinetic and thermodynamic estimation.
• A simple reweighting method can effectively utilize seeding trajectory data.
• The proposed method improves the estimation of both thermodynamic and kinetic quantities.

Conclusions:

• Adaptive sampling strategies, particularly reseeding, require careful implementation and analysis.
• Reweighting techniques are essential for maximizing the utility of trajectory data from enhanced sampling methods.
• This study provides a practical approach to improve the accuracy of MSM-based analyses of biomolecular dynamics.