HelixCorr

HelixCorr predicts correlated mutations within the transmembrane regions of alpha-helical membrane proteins to identify intramolecular amino acid contacts relevant to protein structure and function.

Key Features:

• Target scope: Identifies co-evolving residues specifically within transmembrane regions of alpha-helical membrane proteins.
• Algorithms integrated: Integrates seven distinct prediction algorithms to detect correlated mutations.
• Benchmark dataset: Analyzes correlations across 14 membrane proteins with known three-dimensional structures.
• Consensus predictions: Generates consensus predictions by combining results from multiple algorithms to enhance accuracy and reduce false positives.
• Contact proximity: Reports that 53% of predicted residue pairs lie within one helix turn of an actual helix-helix contact.
• Helix interaction performance: Achieves 83% specificity and 42% sensitivity for predicting interacting helices when using correlations detected by the four most effective algorithms.
• Enrichment at helix-helix contacts: Finds correlated residue pairs at significantly shorter distances than random, particularly at helix-helix contact regions.

Scientific Applications:

• Intramolecular contact identification: Identification of residue pairs that indicate intramolecular amino acid contacts in membrane proteins.
• Helix-helix interaction mapping: Mapping of helix-helix contact regions within alpha-helical membrane proteins.
• Structural modeling support: Providing residue-pair constraints to support construction of structural models of membrane proteins.
• Correlated-mutation analysis: Systematic analysis of co-evolving residues in membrane proteins distinct from soluble protein studies.

Methodology:

Integration of seven distinct prediction algorithms applied to 14 membrane proteins with known three-dimensional structures, generation of consensus predictions by combining multiple algorithm outputs, and evaluation using correlations from the four most effective algorithms.