Calculations of the Absolute Free Energy, Enthalpy, and Entropy of Drug Binding Using Different Potential Models

Area of Science:

• Computational chemistry
• Molecular modeling
• Biophysics

Background:

• Accurate prediction of ligand binding thermodynamics is crucial for drug discovery and molecular design.
• Existing computational models vary in their ability to reproduce experimental binding free energies, enthalpies, and entropies.
• Understanding the sensitivity of these thermodynamic properties to underlying force field parameters is essential for model development.

Purpose of the Study:

• To determine the sensitivity of calculated ligand binding free energy, enthalpy, and entropy to variations in water, protein, and ligand potentials.
• To evaluate the accuracy of different potential models for predicting these thermodynamic properties.
• To identify strategies for improving the accuracy of enthalpy and entropy predictions.

Main Methods:

• Investigated nine protein-ligand systems using four distinct protein targets.
• Employed four different potential models for molecular simulations.
• Assessed the accuracy of calculated free energy, enthalpy, and entropy against experimental data.
• Tested a simple scaling approach for ligand-water and ligand-protein dispersion interactions.

Main Results:

• Several models demonstrated good accuracy in predicting binding free energies.
• Enthalpy and entropy changes were found to be significantly less accurate across most tested models.
• A simple scaling of dispersion interactions led to improved enthalpy and entropy predictions.
• The accuracy of enthalpy and entropy remained a challenge, indicating limitations in current force fields.

Conclusions:

• Current computational models show variable success in predicting ligand binding thermodynamics, with free energy being more reliably reproduced than enthalpy and entropy.
• Simple adjustments to dispersion interactions can partially improve enthalpy and entropy predictions.
• Further advancements necessitate the development of ligand potentials specifically parameterized in conjunction with optimized water models for enhanced accuracy.