Comparison of substructural epitopes in enzyme active sites using self-organizing maps

Area of Science:

• Computational Biology
• Structural Bioinformatics
• Drug Discovery

Background:

• Understanding protein-ligand interactions is crucial for drug discovery.
• Comparing protein binding cavities can reveal functional relationships.
• Existing methods may lack precision in substructural epitope comparison.

Purpose of the Study:

• To develop and present a new algorithm for comparing substructural epitopes in protein binding cavities.
• To enable prediction of potential ligands for unknown protein binding sites.
• To identify functional relationships among proteins based on binding site characteristics.

Main Methods:

• Describing binding-site physicochemical characteristics using pseudocenters of amino acid functional groups.
• Dividing cavities into local regions shaped like pyramids with triangular bases.
• Employing emergent self-organizing maps for clustering similar local regions.
• Superpositioning cavities based on similar local regions to score matches.

Main Results:

• The algorithm correctly identifies similarities between enzymes sharing the same Enzyme Commission (EC) number.
• Enzymes with different EC numbers show no common substructures, indicating functional specificity.
• The method demonstrates potential for accurately comparing protein binding sites.

Conclusions:

• The developed algorithm effectively compares substructural epitopes in protein binding cavities.
• This method can aid in predicting ligand shapes and identifying functional relationships between proteins.
• Further studies are warranted to explore the full potential of this computational approach.