Nonequilibrium Binding Free Energy Simulations: Minimizing Dissipation

Area of Science:

• Computational Chemistry
• Biophysics
• Molecular Dynamics

Background:

• Equilibrium free energy estimation methods are established.
• Nonequilibrium simulations offer parallelization but face convergence challenges.
• Protein-ligand binding free energy estimation is crucial in drug discovery.

Purpose of the Study:

• Investigate challenges in protein-ligand binding free energy estimation via nonequilibrium simulations.
• Analyze the impact of simulation parameters on estimator convergence.
• Provide strategies to enhance simulation efficiency.

Main Methods:

• Nonequilibrium molecular dynamics simulations.
• Physical path sampling for protein-ligand binding.
• Analysis of free energy estimators (e.g., Crooks equation).
• Systematic variation of water models and path collective variables.

Main Results:

• Irreversible work generated during simulations hinders estimator convergence.
• Water models critically affect the convergence rate of free energy estimators.
• Parametrization of path collective variables significantly impacts convergence.
• Identified key factors influencing dissipation in simulations.

Conclusions:

• Nonequilibrium simulations require careful parameter selection for reliable free energy calculations.
• Optimized water models and collective variables enhance convergence speed.
• Practical strategies can minimize dissipation and improve computational efficiency.