GOEGCN

GOEGCN integrates gene regulatory networks and differential expression to perform Gene Ontology (GO) enrichment based on gene coexpression network differences for breast invasive carcinoma (BRCA) subtype classification.

Key Features:

• Integration of Gene Regulatory Networks: Combines biological significance from gene regulatory networks (GRNs) with differential expression analysis to compute weighted differentially expressed genes (weighted DEGs) whose weights reflect a gene's regulatory influence by the number of target genes it regulates.
• Gene Coexpression Network Construction: Constructs coexpression networks for control and experimental groups and compares them to identify significantly different interacting structures.
• GO Enrichment Analysis: Performs Gene Ontology (GO) enrichment analysis that focuses on differences in coexpression network structures to reveal enriched GO terms reflecting functional shifts across BRCA subtypes.
• Binary Classification with Machine Learning: Uses weighted DEGs to train binary classifiers for each breast cancer subtype and evaluates classifier performance using sensitivity, specificity, accuracy, F1 score, and AUC.

Scientific Applications:

• Subtype-specific functional annotation: Reveals enriched GO terms and functional shifts specific to BRCA subtypes.
• Molecular mechanism discovery: Highlights changes in gene interaction mechanisms underlying diversity among BRCA subtypes.
• Therapeutic stratification: Supports development of personalized treatment strategies by providing subtype-specific gene interaction and functional insights.

Methodology:

Integrates differential expression analysis with gene regulatory network information to assign weights to DEGs, constructs and compares coexpression networks for control versus experimental groups to identify significant structural differences, performs GO enrichment based on those network differences, and trains binary classifiers using weighted DEGs with performance evaluated by sensitivity, specificity, accuracy, F1 score, and AUC.