Binding site matching in rational drug design: algorithms and applications

Area of Science:

• Biochemistry
• Computational Biology
• Pharmacology

Background:

• Protein-ligand interactions are vital for biological functions, occurring in ligand-binding pockets.
• Understanding binding pocket physicochemical properties is crucial for basic biological knowledge and drug development.
• Quantifying pocket similarities involves comparing bound ligands or directly matching binding sites using computational methods.

Purpose of the Study:

• To review and categorize 12 widely used computational tools for matching protein ligand-binding pockets.
• To analyze the algorithms, test sets, and performance of these pocket-matching methods.
• To discuss the pharmacological applications of computational pocket matching in drug discovery.

Main Methods:

• Surveyed 12 computational tools for pocket matching.
• Categorized methods based on alignment algorithms.
• Analyzed algorithms, test sets, and performance metrics for each tool.

Main Results:

• The review categorizes 12 pocket-matching tools into five algorithmic groups.
• Detailed analysis of algorithms, test sets, and performance is provided for each method.
• Pharmacological applications including drug repurposing, polypharmacology, and side effect prediction are discussed.

Conclusions:

• Computational pocket matching is essential for drug discovery.
• Future directions include developing meta-predictors, incorporating protein flexibility, and integrating deep learning.
• Accurate pocket matching enhances drug repurposing, polypharmacology, and side effect analysis.