InterMap3D

InterMap3D predicts co-evolving pairs of amino acids within protein sequences by combining homolog identification, multiple sequence alignment, and structural context from the Protein Data Bank (PDB) with three co-evolutionary algorithms to detect residue-residue relationships relevant to protein interactions and function.

Key Features:

• Sequence alignment and homology search: Starts from a single input protein sequence, identifies homologous sequences, and generates an optimal multiple sequence alignment for downstream analysis.
• Structural modeling: Retrieves the most similar three-dimensional (3D) structure from the Protein Data Bank (PDB) to contextualize residues and accepts externally provided multiple sequence alignments.
• Prediction of co-evolving residues: Applies Row and Column Weighing of Mutual Information, Mutual Information/Entropy, and Dependency to predict co-evolving residue pairs from the alignment.
• Visualization: Generates structural visualizations with predicted co-evolving residues highlighted to support interpretation.

Scientific Applications:

• Protein–protein interaction analysis: Infers residue pairs that may mediate protein–protein interfaces.
• Functional site identification: Identifies residues and residue pairs that may constitute functional or active sites.
• Evolutionary conservation studies: Explores co-evolutionary patterns and conservation across homologous proteins.
• Structural biology, molecular evolution, and drug design: Provides residue-level predictions to inform structural experiments, evolutionary analyses, and therapeutic targeting.

Methodology:

Identify homologous sequences, construct a multiple sequence alignment (or accept an external alignment), retrieve the most similar 3D structure from the PDB, apply Row and Column Weighing of Mutual Information, Mutual Information/Entropy, and Dependency to predict co-evolving residues, and generate 3D visualizations of predicted residues.