MSClustering

MSClustering performs multi-level minimum span clustering of networks from distance matrix inputs to reveal hierarchical and phylogenetic structure in biological datasets within the Cytoscape environment.

Key Features:

• Minimum Span Clustering (MSC): Computes automatic minimum span clustering of a network across multiple characteristic levels.
• MSC tree hierarchical representation: Organizes clustering results into an MSC tree that encodes multi-level network structure.
• Distance matrix input: Operates on a distance matrix as the primary input for clustering computations.
• Phylogenetic case studies: Applied to 63 beta coronaviruses and 197 G protein-coupled receptors (GPCRs) to analyze evolutionary relationships.
• Scalability and performance: Demonstrated clustering of 3481 enzymes and reported experimental comparisons showing superior efficiency and reliability relative to five state-of-the-art methods.
• Cytoscape integration: Runs within the Cytoscape environment for network analysis and representation.

Scientific Applications:

• Phylogenetic analysis: Examining evolutionary relationships among organisms and protein families such as beta coronaviruses and GPCRs (63 and 197 datasets respectively).
• Large-scale enzyme clustering: Clustering and structural analysis of large enzyme datasets, exemplified by 3481 enzymes.
• Comparative benchmarking: Evaluating clustering efficiency and reliability against five state-of-the-art methods.

Methodology:

Takes a distance matrix as input, computes minimum span clustering at multiple characteristic levels, and organizes the results into an MSC tree.