Ligand-Enhanced Negative Images Optimized for Docking Rescoring.

Area of Science:

• Computational Chemistry
• Drug Discovery
• Structural Biology

Background:

• Molecular docking is crucial for drug discovery but default scoring functions often miss active ligands.
• Negative image-based (NIB) rescoring enhances docking enrichment by comparing ligand poses to the protein's inverted cavity.
• Optimizing NIB models with greedy search significantly improves docking rescoring yields.

Purpose of the Study:

• To introduce a hybrid approach, LBR-NiB, integrating actual ligand 3D coordinates into NIB models.
• To enhance the accuracy and effectiveness of virtual screening through improved docking rescoring.
• To leverage protein-bound ligand information for more precise pharmacophore modeling.

Main Methods:

• Developed ligand-enhanced brute-force negative image-based optimization (LBR-NiB).
• Incorporated actual ligand 3D coordinates into NIB models for optimization.
• Benchmarked LBR-NiB against previous methods using proinflammatory targets.

Main Results:

• LBR-NiB consistently improves docking enrichment compared to prior R-NiB iterations.
• The enhancement is significant when ligand data provides crucial binding information absent in cavity-based NIB models.
• High-quality ligand data (e.g., X-ray crystallography) and even solvent molecules can improve NIB models.

Conclusions:

• LBR-NiB effectively enhances virtual screening by refining negative image features with protein-bound ligand data.
• This hybrid approach combines the comprehensive nature of NIB models with specific atomic arrangements of ligands.
• The study validates the use of protein-bound ligands to improve docking rescoring for drug discovery.