Predicting protein-ligand binding modes for CELPP and GC3: workflows and insight.

Area of Science:

• Computational chemistry and molecular modeling
• Drug discovery and development

Background:

• The Drug Design Data Resource (D3R) provides crucial benchmarking datasets for advancing computational drug discovery methods.
• Accurate prediction of protein-ligand binding modes is essential for efficient drug design.

Purpose of the Study:

• To integrate, automate, and systematically evaluate developed protein-ligand binding mode prediction methods.
• To assess the impact of using appropriate protein structures on docking and binding mode prediction accuracy.

Main Methods:

• Integration and automation of previously developed protein-ligand binding mode prediction methods.
• Systematic testing using large-scale data from Continuous Evaluation of Ligand Pose Prediction (CELPP) and Grand Challenge 3 (GC3).
• Comparison of template-based and molecular docking methods for binding mode prediction.

Main Results:

• Increasing availability of protein-ligand complex structures in the Protein Data Bank benefits molecular docking methods.
• Selection of an appropriate protein structure for docking significantly enhances binding mode prediction success rates.
• Comparative analysis of template-based and docking method performance.

Conclusions:

• Current computational docking methods show improved performance with larger structural datasets.
• The choice of protein structure is a critical factor for successful binding mode prediction.
• Future research will focus on combining template-based and docking approaches for enhanced prediction accuracy.