eCOMPASS

eCOMPASS evaluates multiple sequence alignments (MSAs) of proteins using a direct coupling analysis (DCA)-based statistical measure to quantify alignment quality and detect biologically relevant residue–residue couplings reflected in 3D protein contacts.

Key Features:

• DCA-based scoring: Uses direct coupling analysis (DCA) to generate scores for residue–residue couplings within MSAs.
• Statistical significance testing: Calculates the statistical significance of congruence between high-scoring directly coupled pairs and corresponding 3D contacts in protein structures.
• Alignment quality metrics: Quantifies alignment quality through DCA-derived metrics to distinguish true evolutionary couplings from random noise.
• Large MSA applicability: Designed to assess large and precise MSAs typical in proteomics studies.
• Validation datasets: Demonstrated on both simulated and real-world MSAs.
• Implementation: Implemented in C++.

Scientific Applications:

• MSA validation against structures: Assess whether MSAs align homologous residues consistent with 3D protein contacts.
• Detection of coevolving residues: Identify compensatory substitutions and coevolving residue pairs indicative of structural or functional constraints.
• Comparative evaluation of alignments: Compare the relative quality of alternative protein MSAs.
• Large-scale proteomics analyses: Support construction and validation of large-scale MSAs to detect subtle biologically relevant patterns.

Methodology:

eCOMPASS employs direct coupling analysis (DCA) to score residue pairs, calculates the statistical significance of congruence between high-scoring coupled pairs and 3D protein contacts, and quantifies alignment quality using DCA-derived metrics; the software is implemented in C++.