Detecting conserved protein complexes using a dividing-and-matching algorithm and unequally lenient criteria for...

Area of Science:

• Computational Biology
• Bioinformatics
• Systems Biology

Background:

• Protein-protein interaction (PPI) data is crucial for understanding biological evolution through conserved subnetworks.
• Existing local alignment methods struggle with strict topological similarity and fail to account for network differences.

Purpose of the Study:

• To propose UEDAMAlign, a new dividing-and-matching method for detecting conserved protein complexes across species.
• To address limitations of current methods by employing unequally lenient criteria for network alignment.

Main Methods:

• UEDAMAlign divides one PPI network into subnetworks and maps proteins to homologous proteins in another network.
• It applies differential leniency criteria to identify common connected components between the two networks.
• Network alignments were performed between S. cerevisiae and D. melanogaster, and H. sapiens and D. melanogaster.

Main Results:

• UEDAMAlign demonstrated superior performance compared to six existing methods in recovering conserved protein complexes.
• The identified complexes showed good agreement with known protein complexes and functional similarity.
• The method effectively handles differences in network topology and size between species.

Conclusions:

• UEDAMAlign is an effective tool for identifying evolutionarily conserved protein complexes.
• The method's ability to use unequally lenient criteria enhances the accuracy of cross-species network alignment.
• This approach advances the study of biological evolution and protein complex conservation.