Identification of protein binding surfaces using surface triplet propensities

Area of Science:

• Computational biology
• Structural bioinformatics
• Drug discovery

Background:

• Predicting protein-protein and protein-ligand interactions is crucial for drug discovery and understanding protein function.
• Existing methods often rely on identifying surface cavities, which may not capture all potential binding sites.
• A novel approach is needed to identify ligand recognition sites more effectively.

Purpose of the Study:

• To develop and evaluate a new algorithm for predicting protein-ligand binding sites.
• To utilize surface atomic group triplets, rather than cavities, for interaction prediction.
• To assess the algorithm's performance against established methods.

Main Methods:

• Developed the surface triplet propensities (STP) algorithm, identifying 455 types of adjacent surface atomic group triplets.
• Trained the algorithm using 309 protein-ligand X-ray structures to determine interface propensities for triplets.
• Validated performance using Q-SiteFinder's dataset and analyzed triplet types related to ligand binding propensity.

Main Results:

• The STP algorithm achieved an 88% success rate in locating protein-ligand binding sites.
• Performance favorably compared to existing grid-based and energy-based approaches.
• Identified a correlation between higher ligand binding propensity and more polarizable surfaces, estimating statistical free energies of interaction.

Conclusions:

• The surface triplet propensities (STP) algorithm offers a robust and accurate method for predicting protein-ligand binding sites.
• This approach provides a valuable tool for identifying druggable binding sites and advancing drug discovery efforts.
• The method is freely available online, facilitating broader application in the scientific community.