GeDiPNet

GeDiPNet integrates curated and text-mined gene–disease associations from DisGeNET, ClinGen, ClinVar, HPO, OrphaNet and PsyGeNET into a harmonized, non-redundant knowledge base to enable statistical comorbidity analysis and polypharmacological target prediction.

Key Features:

• Source integration: Unifies curated associations from DisGeNET, ClinGen, ClinVar, HPO, OrphaNet and PsyGeNET into a single resource.
• Harmonization and standardization: Applies systematic harmonization and manual review to resolve redundancy and inconsistent disease nomenclature, producing a standardized, non-redundant knowledge base.
• Gene coverage: Contains curated disease associations for 12,270 genes and text-mined associations for an additional 20,013 genes.
• Statistical enrichment-based comorbidity prediction: Identifies statistically enriched disease co-occurrences in a hypothesis-free manner based on shared genetic etiology.
• Functional pathway overlap analysis: Assesses comorbidity through overlap of functional pathways between disease-associated gene sets.
• Polypharmacological target prediction: Identifies genes shared among comorbid disease sets as candidate multitarget therapeutic targets.
• Drug–target mapping and benchmarking: Maps predicted targets to drug–target relationships and benchmarks predictions against known multitarget drugs (examples: PDE5A for hypertension/ED, ADA for inflammatory disease, PIK3CA for diabetes/PCOS).
• Case-study outputs: Reports disease-specific findings such as enriched comorbidities for metabolic syndrome (MetS) including obesity, insulin resistance and hypercholesterolemia and associations related to bilirubin metabolism and oxidative stress.
• Putative target identification for MetS: Identified 32 putative targets for MetS component overlap, including validated proteins INSR, JAK2, REN, PPARG and novel candidates DECR1 and HPGDS.

Scientific Applications:

• Comorbidity analysis: Prioritizes disease pairs exhibiting comorbidity risk beyond random expectation by detecting statistically enriched shared genetic etiology and pathway overlap.
• Multitarget drug discovery: Generates candidate polypharmacological targets by identifying genes common to comorbid disease sets and linking them to drug–target relationships.
• Therapeutic hypothesis generation and benchmarking: Produces hypotheses for therapeutic targets and validates inference quality by benchmarking against known multitarget drugs.
• Disease-specific investigation: Supports focused analyses such as metabolic syndrome component interactions and mechanistic links involving bilirubin metabolism and oxidative stress.

Methodology:

Systematic harmonization and manual review of associations from DisGeNET, ClinGen, ClinVar, HPO, OrphaNet and PsyGeNET to produce a standardized, non-redundant knowledge base; statistical enrichment models for hypothesis-free identification of enriched disease co-occurrences based on shared genetic etiology and functional pathway overlap; mapping of shared genes to drug–target relationships; benchmarking against known multitarget drugs.