Simulating rare events using a weighted ensemble-based string method.

Area of Science:

• Computational Chemistry
• Statistical Mechanics
• Biophysics

Background:

• Simulating molecular dynamics in high-dimensional phase spaces is computationally challenging.
• Identifying key reaction pathways is crucial for understanding molecular mechanisms.
• Existing path sampling methods may struggle with efficiency and accuracy in complex systems.

Purpose of the Study:

• To introduce an extension of the weighted ensemble (WE) path sampling method.
• To enable efficient sampling of one-dimensional paths in high-dimensional phase spaces.
• To accurately identify principal reaction pathways for equilibrium and non-equilibrium systems.

Main Methods:

• The method extends the weighted ensemble (WE) path sampling approach.
• It incorporates principles from the finite-temperature string method.
• Adaptive refinement of Voronoi tessellations in order parameter space is employed.

Main Results:

• The method efficiently samples systems, guiding adaptive refinement of order parameter space.
• Generating points of the tessellation converge to the principal reaction pathway.
• Demonstrated application on 2D Brownian motion models and protein conformational changes.

Conclusions:

• The extended WE method provides efficient sampling for complex molecular systems.
• It accurately identifies principal reaction pathways, crucial for mechanistic studies.
• Applicable to diverse systems, from simple models to protein dynamics.