Communication: Microsecond peptide dynamics from nanosecond trajectories: a Langevin approach

Area of Science:

• Computational Chemistry
• Statistical Mechanics
• Biophysics

Background:

• Time series analysis is crucial for understanding complex system dynamics.
• Traditional methods often require extensive simulation data or continuous trajectories.
• Predicting long-time dynamics from short trajectories remains a challenge.

Purpose of the Study:

• To introduce and validate the data-driven Langevin equation (dLE) for modeling complex dynamics.
• To assess the dLE's ability to reproduce statistical and dynamical features from limited data.
• To demonstrate the dLE's applicability in predicting biomolecular system dynamics.

Main Methods:

• Estimating drift and diffusion fields from time series data using the dLE.
• Utilizing local information for dLE propagation, not requiring Boltzmann weighting or continuous trajectories.
• Applying dLE to extensive molecular dynamics simulations of peptide helix unfolding/refolding.

Main Results:

• The dLE successfully reproduces essential statistical and dynamical features of the original time series.
• The approach can predict microsecond conformational dynamics from only a few hundred nanosecond trajectories.
• Quantitative reproduction of the free energy landscape and conformational dynamics was achieved.

Conclusions:

• The dLE offers a promising approach for predicting long-time dynamics from short, parallel trajectories.
• Its applicability is primarily limited by the initial conformational space sampling.
• The dLE effectively models microsecond dynamics in biomolecular systems, validated by peptide helix simulations.