NET-GE

NET-GE performs network-based enrichment analysis of gene and protein sets by extracting STRING interactome sub-networks that connect proteins sharing Gene Ontology (GO) Biological Process annotations to identify and extend functional annotations related to phenotypes.

Key Features:

• Network-based enrichment: Incorporates physical and functional relationships from biological interaction networks to enrich gene/protein set analysis.
• STRING database utilization: Exploits the STRING interactome to build minimal connecting networks for proteins annotated with identical GO terms.
• GO Biological Process modules: Extracts sub-networks that connect proteins sharing GO terms specifically related to Biological Processes (BP).
• Annotation mapping and collection: Maps input protein sets onto GO-specific sub-networks and collects corresponding annotations.
• Enhanced annotation discovery (benchmark): In a benchmark of 244 protein sets associated with Mendelian diseases, extracted novel GO terms in 58% of cases missed by standard methods and identified new annotations not present in the input set for 35% of cases.
• Small dataset efficacy: Performs enrichment analysis with a minimum of two proteins, enabling application to small protein sets.
• Comparative case analyses: Provides insights from analyses of six specific cases that demonstrate improved biological interpretation relative to other network-based methods.

Scientific Applications:

• Annotation expansion and variant interpretation: Expands sets of potentially related genes/proteins and aids interpretation and prioritization of newly identified variants, particularly in the context of Mendelian diseases.
• Functional module definition and mechanism discovery: Defines protein functions within interaction networks via GO-specific modules to reveal molecular mechanisms underlying phenotypes.

Methodology:

Uses the STRING interactome to extract minimal sub-networks connecting proteins that share GO Biological Process terms and maps input protein sets onto these sub-networks to collect and extend annotations.