ConsCapra07

ConsCapra07 evaluates residue conservation at specific sites in multiple sequence alignments (MSAs) using an information-theoretic Jensen-Shannon divergence framework to predict functionally important protein residues.

Key Features:

• Jensen-Shannon Divergence: Uses Jensen-Shannon divergence as the core metric to quantify position-specific sequence conservation.
• Sequential Neighbor Consideration: Incorporates a heuristic that accounts for the estimated conservation of sequentially neighboring sites to refine conservation estimates.
• Performance Superiority: Demonstrated superior performance relative to Shannon entropy and other conservation-based measures for identifying catalytic sites and residues near bound ligands, with reduced efficacy for protein-protein interface residues.
• Data Accessibility: Provides access to datasets and code for the evaluated conservation measures.

Scientific Applications:

• Functional residue prediction: Predicts functionally important residues from MSAs, including catalytic and ligand-binding residues.
• Protein structure and function inference: Highlights conserved positions to inform analyses of protein stability, enzymatic activity, and ligand interactions.
• Evolutionary analysis and annotation: Supports studies of evolutionary relationships and functional annotation of newly sequenced proteins by integrating neighbor conservation context.
• Protein-protein interface detection (limitation): Is less effective for identifying residues involved in protein-protein interfaces and therefore is not recommended as a primary method for that application.

Methodology:

Computes Jensen-Shannon divergence on positions in a multiple sequence alignment (MSA) and applies a heuristic incorporating the estimated conservation of sequential neighbors, with performance evaluated against Shannon entropy and other conservation measures.