AutoFlex-Dock: New Molecular Docking Workshop Supports Deciphering Protein-Ligand Interactions

Area of Science:

• Structural bioinformatics
• Computational drug discovery
• Molecular modeling

Background:

• Protein-ligand interactions (PLIs) are fundamental to understanding molecular mechanisms in biology and medicine.
• Molecular docking is a key technique for predicting PLIs, but its accuracy is often limited by ligand flexibility.
• Accurate prediction of PLIs is crucial for rational drug design and discovery.

Purpose of the Study:

• To develop and validate a novel strategy to improve the accuracy of protein-ligand interaction prediction.
• To address the limitations of ligand flexibility in molecular docking.
• To create a user-friendly computational tool for exploring PLIs.

Main Methods:

• A number of torsion bond (NTB)-based strategy was proposed to enhance ligand sampling in molecular docking.
• The NTB-based strategy was integrated into the AutoDock program.
• A web server, AutoFlex-Dock, was developed, incorporating NTB strategy, binding free energy calculations, and multiple binding pose analysis.

Main Results:

• The NTB-based strategy achieved a sampling success rate of 62.8% at a 1.0 Å RMSD threshold.
• This represents an 8–21% improvement compared to single search algorithms in AutoDock.
• The AutoFlex-Dock server provides a user-friendly platform for PLI analysis.

Conclusions:

• The NTB-based strategy significantly improves the accuracy of protein-ligand interaction predictions.
• The AutoFlex-Dock server offers a valuable and convenient tool for researchers in structural bioinformatics and drug discovery.
• This work facilitates a deeper understanding of molecular mechanisms underlying PLIs.