On Calculating Free Energy Differences Using Ensembles of Transition Paths

Area of Science:

• Computational chemistry
• Molecular dynamics
• Drug discovery

Background:

• Binding free energy and kinetics (association and dissociation rates) are crucial for drug design.
• Calculating binding kinetics from simulations is challenging due to state definitions and simulation details.
• Existing methods for calculating free energies from rate ratios lack corrections for practical simulation aspects.

Purpose of the Study:

• To derive and apply correction terms for calculating binding free energies from molecular simulation rate constants.
• To improve the accuracy of free energy calculations from simulation data, particularly when using rate ratios.

Main Methods:

• Derivation of correction terms for binding free energy calculations from reactive trajectories.
• Application of these corrections to a host-guest system using weighted ensemble simulations.
• Comparison of corrected free energy calculations with alchemical free energy perturbation and experimental data.

Main Results:

• The derived correction terms significantly reduced the error in binding free energy calculations from 3.4 to 0.76 kcal/mol.
• Corrections improved the consistency between free energies calculated from rate ratios and alchemical methods.
• The study revisited a host-guest system from a blind prediction challenge, resolving previous deviations.

Conclusions:

• The developed correction terms enhance the accuracy of binding free energy calculations from simulation rate constants.
• These corrections are valuable for computational drug design, improving the reliability of predicted drug efficacy.
• The methodology is broadly applicable to various simulation-based free energy calculation techniques.