Combining stochastic resetting with Metadynamics to speed-up molecular dynamics simulations

Area of Science:

• Computational chemistry and biophysics
• Molecular dynamics simulations
• Enhanced sampling techniques

Background:

• Metadynamics enhances molecular dynamics simulations but requires identifying suitable collective variables (CVs).
• Finding optimal CVs is often challenging and not known a priori.
• Stochastic resetting is a CV-free enhanced sampling method.

Purpose of the Study:

• To combine Metadynamics with stochastic resetting for enhanced molecular dynamics simulations.
• To investigate the efficiency of the combined approach compared to individual methods.
• To develop a method for extracting unbiased results from resetting Metadynamics.

Main Methods:

• Integration of stochastic resetting into Metadynamics simulations.
• Performance evaluation using suboptimal and optimal collective variables.
• Development of a new method for calculating unbiased mean first-passage times.

Main Results:

• The combined Metadynamics and stochastic resetting approach yields greater acceleration than either method alone.
• Resetting Metadynamics with suboptimal CVs achieves speedups comparable to optimal CVs.
• The proposed method improves the speedup-accuracy tradeoff in Metadynamics simulations with resetting.

Conclusions:

• Stochastic resetting is a viable alternative to optimizing CVs in Metadynamics, offering significant speedups with minimal computational cost.
• The combined method accelerates a wide range of molecular simulations.
• This work provides a more efficient and accurate approach to enhanced sampling in molecular dynamics.