SE-DMTG

SE-DMTG detects protein complexes within Protein-Protein Interaction Networks (PPINs) by integrating topological structure and Gene Ontology (GO) annotations and applying a seed-extended density and modularity-based fitness function to capture complexes of varying densities and modularities.

Key Features:

• Integration of Topological and Functional Data: Constructs a weighted PPIN by integrating topological information (common neighbors) and Gene Ontology (GO) annotations to reflect interaction strength and functional similarity.
• Seed Selection Strategy: Selects seed nodes using common neighbors and GO annotations to prioritize nodes with high connectivity and functional relevance.
• Fitness Function for Complex Detection: Employs a fitness function designed to detect protein complexes across diverse densities and modularities.
• Weighted PPIN Construction: Weights network edges based on combined topological measures and GO-based functional similarity.

Scientific Applications:

• Disease mechanism and drug discovery: Improves detection accuracy of protein complexes to support studies of complex diseases and target identification for drug discovery.
• Cross-species PPIN analysis: Handles varying densities and modularities, making it adaptable to PPINs from multiple organisms.
• Benchmarking and functional enrichment: Demonstrates superior performance on yeast PPINs measured by F-measure and Jaccard index and achieves improved functional enrichment.

Methodology:

Constructs a weighted PPIN by integrating common neighbors and GO annotations; selects seed nodes based on connectivity and GO-derived functional relevance; evaluates and detects complexes using a fitness function that accounts for density and modularity.