Force-momentum-based self-guided Langevin dynamics: a rapid sampling method that approaches the canonical ensemble

Area of Science:

• Computational Chemistry
• Molecular Dynamics
• Statistical Mechanics

Background:

• Self-guided Langevin dynamics (SGLD) accelerates conformational searching by frequency-selective motion enhancement/suppression.
• Recent advances allow SGLD for accurate quantitative studies via its partition function for reweighting to canonical ensemble properties.

Purpose of the Study:

• To present a novel force-momentum-based self-guided Langevin dynamics (SGLDfp) method for direct canonical ensemble sampling.
• To evaluate SGLDfp's ability to maintain canonical distribution and accelerate conformational searching.

Main Methods:

• Developed SGLDfp by incorporating interaction forces into the guiding force to counteract momentum-based perturbations.
• Tested SGLDfp on various systems to assess its sampling accuracy and efficiency.
• Compared SGLDfp and SGLD performance in crossing energy barriers.

Main Results:

• SGLDfp simulations approximately maintain the canonical ensemble distribution.
• SGLDfp significantly accelerates conformational searching, enabling crossing of high energy barriers (>15 kT).
• SGLDfp is size-extensive and effective for large systems.

Conclusions:

• SGLDfp is an efficient method for conformational searching and sampling, particularly for studies requiring preservation of accessible conformational space.
• The method is suitable for applications like free energy calculations and protein folding simulations.
• SGLDfp offers a powerful tool for exploring complex molecular landscapes.