CovET: A covariation-evolutionary trace method that identifies protein structure-function modules.

Area of Science:

• Computational Biology
• Protein Engineering
• Evolutionary Biology

Background:

• Understanding residue interactions is crucial for protein design and interpreting genetic variants.
• Existing methods often overlook phylogenetic divergence, limiting insights into functional impacts.

Purpose of the Study:

• To develop a novel covariation analysis method (CovET) that incorporates phylogenetic divergences.
• To assess CovET's performance in identifying functionally critical residue pairs and protein structures.

Main Methods:

• Reframing covariation analysis within the Evolutionary Trace framework.
• Systematically penalizing covariation patterns inconsistent with evolutionary coupling at each divergence event.
• Evaluating CovET's ability to predict structural contacts, clusters, ligand binding sites, and correlate with epistasis data.

Main Results:

• CovET matches existing methods for predicting individual structural contacts.
• CovET significantly outperforms other methods in identifying structural clusters of coupled residues and ligand binding sites.
• CovET accurately recovered the allosteric activation pathway in the dopamine D2 receptor.

Conclusions:

• CovET effectively identifies functionally critical residue pairs by accounting for phylogenetic divergences.
• The method enhances the understanding of protein structure-function relationships, epistasis, and allosteric mechanisms.
• CovET offers a valuable complement to existing methods for protein design and functional interpretation.