Combining Iteration-Free Polarization with Large Time Step Stochastic-Isokinetic Integration

Area of Science:

• Computational Chemistry
• Molecular Dynamics Simulations
• Physical Chemistry

Background:

• Accelerating molecular dynamics (MD) simulations is crucial for studying complex condensed phase systems.
• Polarizable force fields offer improved accuracy but are computationally expensive.
• Existing methods often require computationally intensive self-consistent field (SCF) steps or short time steps.

Purpose of the Study:

• To develop and validate a novel computational approach for accelerating MD simulations using polarizable force fields.
• To combine an extended Lagrangian method that eliminates the SCF step (iEL/0-SCF) with a stochastic integration scheme (SIN(R)) for long time steps.
• To assess the performance of the combined method across various condensed phase systems.

Main Methods:

• Implementation of a new extended Lagrangian approach (iEL/0-SCF) to bypass the SCF calculation.
• Integration of the iEL/0-SCF method with a stochastic resonance (SIN(R)) multiple time scale algorithm for extended time steps.
• Testing the combined scheme on diverse systems: bulk water, concentrated salt solution, aqueous peptide, and solvated protein.

Main Results:

• The combined iEL/0-SCF and SIN(R) approach demonstrates robust performance across different condensed phase systems.
• Computational speed-ups of approximately 6-8 times were achieved compared to standard MD simulations with a traditional SCF solver.
• The method effectively handles many-body potential energy surfaces in equilibrium simulations.

Conclusions:

• The combination of iEL/0-SCF and SIN(R) represents a significant advancement for accelerating MD simulations with polarizable force fields.
• This integrated approach offers substantial computational gains beyond what either method can achieve individually.
• The developed method enables more efficient equilibrium simulations of condensed phase systems.