Conformational Sampling Reveals Amino Acids with a Steric Influence on Specificity

Area of Science:

• Computational biology
• Structural bioinformatics
• Molecular modeling

Background:

• Protein structure comparison is crucial for identifying homologous proteins and understanding function.
• Conformational flexibility, even subtle, can significantly alter the apparent similarity of protein structures, particularly binding sites.
• Existing methods may struggle to accurately compare structures with localized conformational variations.

Purpose of the Study:

• To introduce Flexible Aggregate Volumetric Analysis (FAVA), an algorithm designed to compare ligand binding sites while accounting for subtle, localized flexibility.
• To develop a method that integrates multiple conformational samples to represent binding site geometry.
• To demonstrate FAVA's ability to mitigate comparison errors caused by conformational variations.

Main Methods:

• FAVA integrates hundreds of conformational samples from molecular simulations.
• The algorithm characterizes ligand binding sites by their frequent geometric appearance, excluding rare conformations.
• Analysis involves comparing serine protease and enolase families with known binding preferences.

Main Results:

• FAVA successfully classified protein families with different binding preferences despite substantial binding site variations.
• The algorithm demonstrated the ability to identify specific amino acids influencing ligand binding specificity by examining individual amino acid motion.
• FAVA effectively mitigated comparison errors arising from small conformational changes.

Conclusions:

• FAVA provides a robust method for comparing protein structures, particularly ligand binding sites, by effectively handling conformational flexibility.
• The algorithm enhances the accuracy of remote homology detection and functional site analysis.
• FAVA's ability to analyze conformational ensembles offers new insights into protein dynamics and ligand interactions.