Clustering by common friends finds locally significant proteins mediating modules

Area of Science:

• Systems Biology
• Bioinformatics

Background:

• Protein interaction networks are extensively studied, focusing on topology and modules.
• Identifying proteins that mediate module functions within these networks remains a challenge.

Purpose of the Study:

• To develop and apply a novel algorithm for identifying locally significant proteins (mediators) in large-scale protein interaction networks.
• To investigate the hierarchical organization and functional significance of these mediators and their associated clusters.

Main Methods:

• A layered clustering algorithm was employed to group proteins based on neighborhood similarity.
• Locally significant proteins, termed mediators, were identified as key connectors between clusters.
• The algorithm was applied to analyze the protein interaction network of yeast.

Main Results:

• Identified hierarchical organization of clusters and mediators, with clusters acting as mediators for others.
• Achieved 71% precision and 61% recall when comparing identified clusters to known yeast complexes.
• Mediators were enriched in diverse processes, particularly biomolecule translocation, and clusters were enriched in functions.

Conclusions:

• Mediators play a crucial role in linking protein clusters and are highly conserved across genomes.
• The findings provide insights into the functional organization of protein interaction networks, exemplified by membrane protein translocation.