Identification of functional modules in a PPI network by bounded diameter clustering.

Area of Science:

• Bioinformatics
• Computational Biology
• Graph Theory

Background:

• Protein-protein interaction (PPI) networks are crucial for understanding protein function.
• Identifying dense subgraphs (functional modules) is key to inferring protein behavior.
• Existing methods for biological complex identification face challenges with large PPI datasets.

Purpose of the Study:

• To develop a novel, efficient graph theoretic clustering algorithm for detecting functional modules in large PPI graphs.
• To present a method based on finding bounded diameter subgraphs around a seed node.
• To compare the algorithm's performance against established methods like MCL, Core-Attachment, and MCODE.

Main Methods:

• A new graph theoretic clustering algorithm is proposed.
• The algorithm identifies densely connected regions by finding bounded diameter subgraphs.
• The method was tested on the yeast PPI graph.

Main Results:

• The algorithm demonstrated simplicity and efficiency compared to other graph clustering methods.
• Performance was evaluated against MCL, Core-Attachment, and MCODE algorithms on yeast PPI data.
• The results indicate the algorithm's potential for accurate biological complex identification.

Conclusions:

• The developed algorithm offers a simple and efficient approach to identifying functional modules in PPI networks.
• This method contributes to faster and more accurate biological complex discovery.
• The findings suggest a promising new tool for analyzing large-scale protein interaction data.