CB-Dock: a web server for cavity detection-guided protein-ligand blind docking.

Area of Science:

• Computational biology
• Structural bioinformatics
• Drug discovery

Background:

• The rapid increase in protein structure data necessitates efficient methods for exploiting structural information.
• Predicting protein-ligand binding sites and affinity is crucial for computer-aided drug discovery.
• Existing docking tools often require pre-defined binding sites, limiting automated prediction.

Purpose of the Study:

• To develop a user-friendly blind docking web server for automated prediction of protein binding sites and modes.
• To introduce a novel curvature-based cavity detection approach for identifying binding pockets.
• To provide an efficient tool for drug discovery research.

Main Methods:

• Development of the CB-Dock web server.
• Implementation of a curvature-based cavity detection algorithm to predict binding sites.
• Integration of Autodock Vina for docking simulations.
• Benchmarking against state-of-the-art blind docking tools.

Main Results:

• CB-Dock successfully predicts binding sites and modes without prior user input.
• The curvature-based method accurately calculates binding site centers and sizes.
• Achieved a ~70% success rate for top-ranking poses within 2 Å RMSD of X-ray poses.
• Outperformed other blind docking tools in benchmark tests.

Conclusions:

• CB-Dock provides an effective solution for automated blind docking.
• The novel cavity detection approach enhances binding site prediction accuracy.
• CB-Dock is a valuable, freely available resource for the drug discovery community.