MSClustering: A Cytoscape Tool for Multi-Level Clustering of Biological Networks

Area of Science:

• Computational Biology
• Bioinformatics
• Network Analysis

Background:

• Complex networks are prevalent in biological systems.
• Analyzing these networks requires efficient visualization and clustering tools.
• Existing methods may lack efficiency or reliability for large datasets.

Purpose of the Study:

• Introduce MSClustering, an efficient software package for network analysis in Cytoscape.
• Demonstrate the application of MSClustering for phylogenetic analysis and studying evolutionary relationships.
• Evaluate the performance of MSClustering against state-of-the-art clustering methods.

Main Methods:

• MSClustering utilizes a distance matrix as input to perform minimum span clustering (MSC).
• The software visualizes multi-level clustering results as an MSC tree for overall network structure.
• Phylogenetic analysis was performed on datasets including beta coronaviruses and GPCRs.

Main Results:

• MSClustering successfully clustered large biological systems, including 3481 enzymes.
• Demonstrated phylogenetic applications in analyzing evolutionary relationships of 63 beta coronaviruses and 197 GPCRs.
• Experimental comparisons showed MSClustering superior to five other methods in clustering efficiency and reliability.

Conclusions:

• MSClustering is an efficient and reliable tool for visualizing and analyzing complex biological networks.
• The software offers valuable phylogenetic insights into evolutionary relationships.
• MSClustering provides a robust solution for large-scale network clustering problems.