CoeViz: a web-based tool for coevolution analysis of protein residues

Area of Science:

• Bioinformatics
• Computational Biology
• Structural Biology

Background:

• Proteins require specific folded structures for function, with active site residues often distant in sequence.
• Traditional methods struggle to identify functional sites in multi-functional proteins.
• Evolutionary information, particularly coevolving residues, enhances protein structure and function prediction.

Purpose of the Study:

• To introduce CoeViz, a web-based tool for analyzing and visualizing pairwise coevolution of amino acid residues.
• To provide a versatile platform for identifying functional residues and structural domains in proteins.

Main Methods:

• CoeViz computes mutual information, chi-square, Pearson correlation, and joint Shannon entropy for coevolution analysis.
• Includes phylogeny correction, sequence dissimilarity correction, and gap correction for covariance scores.
• Visualizes residue relationships using cluster trees, heat maps, circular diagrams, and 3D structure highlighting.

Main Results:

• CoeViz analyzes pairwise coevolution of amino acid residues using multiple metrics.
• The tool offers enhanced visualization options for residue relationships.
• It can process long, unstructured, and multi-domain proteins, unlike other tools.

Conclusions:

• CoeViz offers a practical resource for sequence-based protein annotation, identifying functional residue clusters and domains.
• It serves as a resource for coevolution matrices for computational biologists developing machine learning models.
• The tool is integrated into the POLYVIEW-2D server.