MatrixMatchMaker

MatrixMatchMaker predicts protein-protein interactions by detecting coevolutionary signals in phylogenetic and evolutionary distance matrices.

Key Features:

• Coevolution Detection: Identifies coevolving protein partners by finding large common submatrices within pairs of evolutionary distance matrices, using submatrix size as an indicator of coevolution strength.
• Flexibility in Analysis: Operates on matrices of varying sizes and scales to accommodate paralogous sequences, detect lineage-specific coevolutionary events, and predict multiple interaction partners.
• Application to Human Genome Data: Has been applied to predict protein-protein interactions within the human genome by prioritizing proteins that exhibit strong coevolutionary signals.
• Insights into Biological Networks: Produces highly connected predicted interaction networks that indicate putative interactions beyond current experimental annotations.

Scientific Applications:

• Systems Biology: Enables construction and refinement of protein interaction networks for modeling cellular pathways and processes based on coevolutionary evidence.
• Functional Genomics: Supports generation of hypotheses about protein function and interaction partners from coevolutionary patterns.
• Evolutionary Studies: Facilitates detection and analysis of lineage-specific coevolutionary events to study historical dynamics of protein interactions.
• Target Identification: Provides candidate protein-protein interactions that can be evaluated as potential targets for therapeutic intervention based on coevolutionary support.

Methodology:

Compares pairs of phylogenetic/evolutionary distance matrices to identify large common submatrices and uses submatrix size as a metric of coevolution strength while supporting matrices of varying sizes to detect lineage-specific coevolution and multiple partners.