Binding Affinity via Docking: Fact and Fiction.

Area of Science:

• Computational chemistry
• Structural biology
• Drug discovery

Background:

• Small molecule docking, introduced by Kuntz et al. in 1982, estimates ligand binding poses and affinities.
• Early docking methods focused on pose prediction, with limited accuracy in binding affinity estimation.
• Despite advancements, docking's reliability for predicting binding affinity remains a significant challenge in computational drug design.

Purpose of the Study:

• To critically evaluate the suitability of small molecule docking for accurate binding affinity prediction.
• To identify key limitations hindering reliable binding affinity predictions using current docking approaches.
• To offer insights and potential solutions for improving docking-based affinity predictions.

Main Methods:

• Review and critical analysis of existing docking methodologies and scoring functions.
• Discussion of fundamental challenges in molecular docking, including solvent effects, hydrogen bonding, and system dynamics.
• Comparative assessment of historical and current accuracy in binding affinity prediction.

Main Results:

• Docking methods, despite extensive development, show limited improvement in binding affinity prediction accuracy over the past 10-20 years.
• Key factors limiting accuracy include inadequate representation of solvent (water) interactions, poor description of hydrogen bonds, and the neglect of system dynamics.
• The inherent limitations suggest docking may not be the optimal tool for precise binding affinity calculations.

Conclusions:

• Reconsideration of critical aspects is necessary for realistic binding affinity prediction via docking.
• Improvements in describing solvent effects, hydrogen bonding, and molecular dynamics are essential.
• Further research and methodological advancements are required to enhance the predictive power of docking for binding affinity in drug discovery.