Calculation of absolute protein-ligand binding affinity using path and endpoint approaches

Area of Science:

• Computational chemistry
• Molecular modeling
• Biophysics

Background:

• Accurate calculation of protein-ligand binding affinities is crucial for drug discovery.
• Existing approximate methods often lack sufficient accuracy and reliability.

Purpose of the Study:

• To compare rigorous and approximate methods for calculating absolute binding affinities.
• To evaluate the performance of umbrella sampling versus endpoint methods for protein-ligand complexes.

Main Methods:

• Rigorous method: Umbrella sampling to determine the potential of mean force.
• Approximate method: Endpoint approaches estimating enthalpy, solvation free energy, and entropy.
• Systems studied: FKBP protein with 4-hydroxy-2-butanone and FK506.

Main Results:

• Umbrella sampling results show good agreement with experimental binding affinities.
• Approximate endpoint approaches exhibit critical shortcomings and variations.
• Conventional entropy estimation procedures yield disparate results for simple systems.

Conclusions:

• Rigorous methods like umbrella sampling are reliable for binding affinity calculations.
• Approximate methods, especially for entropy, require careful validation due to potential inaccuracies.
• This study highlights the importance of choosing appropriate computational methods for accurate molecular binding predictions.