MGclus

MGclus detects modules in large-scale biological interaction networks by identifying strongly interconnected neighborhoods that correspond to functional units for gene and protein annotation.

Key Features:

• Module Detection: Detects modules defined by strongly interconnected neighborhoods in biological interaction networks.
• Performance: Demonstrated superior performance in benchmark tests on random graphs with varying levels of noise and comparable or better performance on biological protein interaction networks.
• Implementation: Implemented in Java and leverages the JGraphT graph library for graph representation and operations.
• Robustness and Scalability: Provides robustness and efficiency in handling large datasets.
• Functional Annotation Support: Identifies modules likely to represent protein complexes or biological pathways for gene and protein annotation.

Scientific Applications:

• Functional annotation of genes and proteins: Aids functional annotation by identifying network modules that correspond to functional units.
• Identification of complexes and pathways: Detects modules that are likely to represent protein complexes or biological pathways.
• Analysis of protein interaction networks: Analyzes large-scale protein interaction networks to investigate modular organization and cellular processes.

Methodology:

Identifies network clusters by detecting neighborhoods with strong internal connectivity and handling varying degrees of noise, and was validated using benchmark tests on random graphs with varying levels of noise.