Improving the efficiency and reliability of free energy perturbation calculations using overlap sampling methods

Area of Science:

• Computational chemistry
• Molecular dynamics simulations
• Statistical mechanics

Background:

• Accurate free energy calculations are crucial for understanding molecular systems.
• Conventional methods often struggle with computational cost and reliability for complex systems.
• Efficiently obtaining precise free energy estimates remains a significant challenge.

Purpose of the Study:

• To introduce and validate the overlap sampling (OS) method for efficient and reliable free energy calculations.
• To assess the performance of OS against conventional free energy perturbation (FEP) techniques.
• To demonstrate the applicability of OS across diverse molecular systems.

Main Methods:

• The overlap sampling (OS) method, which samples overlapping phase space during perturbation calculations.
• Testing OS at two optimization levels: simple OS and full optimization (equivalent to Bennett's method).
• Comparison with direct FEP averaging on three alchemical mutation systems: anion in water, methanol/ethane mutation, and adenosine modification.

Main Results:

• OS techniques significantly improve the reliability of free energy estimates at both optimization levels.
• Bennett's method, when implemented with OS, shows particularly striking performance.
• The OS method accelerates convergence and handles large perturbations, enhancing computational efficiency.

Conclusions:

• The overlap sampling method provides a robust and efficient approach for free energy calculations in molecular simulations.
• OS offers improved accuracy and precision without compromising the simulation process.
• This technique is a valuable general perturbation method for computing free energy differences.