The TEMPO integrator: accelerating molecular simulations by temporally multiscale force prediction

Area of Science:

• Computational biology
• Biophysics
• Molecular modeling

Background:

• Molecular dynamics (MD) simulations are crucial for studying biomolecular processes.
• Current MD methods face computational inefficiencies, limiting their scale and application.
• Enhanced sampling methods often compromise kinetic accuracy or require prior knowledge.

Purpose of the Study:

• To develop a computationally efficient simulation method for biomolecular processes.
• To improve the speed of molecular dynamics simulations without losing kinetic detail.
• To overcome limitations of existing enhanced sampling techniques.

Main Methods:

• Developed the temporally multiscale prediction (TEMPO) Integrator.
• Integrated TEMPO into a multiscale Brownian dynamics (MSBD) simulation tool.
• Predicted forces at progressively larger intervals to reduce force evaluations.

Main Results:

• Achieved 27- to 32-fold efficiency gains for intrinsically disordered protein models.
• Demonstrated a seven-fold efficiency improvement for nucleocytoplasmic transport simulations.
• MSBD preserved kinetic properties like reaction rates, unlike conventional methods.

Conclusions:

• The MSBD tool with the TEMPO integrator offers rapid and accurate molecular simulations.
• This approach leverages energy landscape multiscale structure for efficiency.
• TEMPO is generalizable and can complement existing enhanced sampling methods.