Recent theoretical and computational advances for modeling protein-ligand binding affinities

Area of Science:

• Computational chemistry
• Biophysics
• Molecular modeling

Background:

• Accurate prediction of protein-ligand binding free energies is crucial for drug discovery.
• Existing computational methods vary in complexity and applicability.
• Understanding the theoretical basis of these methods is essential for their effective use.

Purpose of the Study:

• To review recent theoretical and algorithmic advances in modeling protein-ligand binding free energies.
• To describe the statistical mechanics theory of noncovalent association.
• To review and compare major computational models and algorithms.

Main Methods:

• Free energy perturbation (FEP)
• Double decoupling (DD)
• Binding Energy Distribution Analysis Method (BEDAM)
• Potential of Mean Force (PMF)
• Mining Minima
• Molecular Mechanics with the Poisson-Boltzmann or Surface Area continuum solvation model (MM/PBSA)

Main Results:

• The fundamental formulas for computational methods are derived from statistical mechanics.
• Various computational models share common theoretical origins despite differing features and limitations.
• Emphasis is placed on modeling conformational reorganization and entropic effects.

Conclusions:

• A unified theoretical framework underlies diverse computational approaches for binding free energy calculations.
• The choice of method depends on specific application requirements and limitations.
• Continued development in theoretical and algorithmic approaches is advancing the field.