Drug Repurposing: From Data to Evidence

SgTxGNN is a drug repurposing prediction platform based on Harvard's TxGNN model. We use AI to predict 31,543 drug repurposing candidates and provide evidence-based validation reports for 745 Singapore HSA-approved medications.

Not just "possibly effective" — we show you where the evidence is. L1-L5 evidence classification helps researchers quickly assess prediction credibility.

Browse High Evidence Drugs Learn Methodology

Drug Search

Enter a drug name or disease name to search for repurposing possibilities and clinical evidence. Supports generic names, brand names, and disease keywords.

What Makes Us Different

Most drug repurposing tools only provide "possibly effective" scores. SgTxGNN integrates ClinicalTrials.gov, PubMed literature, drug interactions, and more — with L1-L5 evidence classification to show which predictions are worth pursuing.

From Prediction to Evidence
Each report integrates clinical trial IDs (NCT), literature indices (PMID), and Singapore HSA license information with complete evidence traceability.

Five-Level Evidence Classification
L1 (Multiple Phase 3 RCTs) to L5 (Model prediction only) with Go / Proceed / Hold recommendations for quick candidate screening.

Singapore HSA Coverage
Focused on HSA-registered medications covering 745 drugs and 4,589 predicted indications. Reports include Singapore license status.

Dual Validation (KG+DL)
1,217 predictions validated by both Knowledge Graph and Deep Learning methods, providing higher confidence for prioritization.

Key Statistics

745

Drugs Analyzed

31,543

Repurposing Candidates

4,589

Diseases Covered

1,217

Dual Validated (KG+DL)

Prediction Sources

Source	Count	Description
KG+DL	1,217	Both Knowledge Graph and Deep Learning agree — highest confidence
DL	27,883	Deep Learning predictions with confidence scores
KG	2,443	Knowledge Graph predictions based on biological relationships

Category	Description	Link
High Evidence	L1-L2, prioritize for evaluation	View Drugs
Medium Evidence	L3-L4, needs additional evidence	View Drugs
Predictions Only	L5, research direction reference	View Drugs
Full Drug List	All 745 drugs (searchable)	Drug List
Methodology	How predictions are made	Methodology
Downloads	CSV / JSON formats	Downloads

About This Project

This system uses Harvard Zitnik Lab's TxGNN deep learning model published in Nature Medicine to predict potential new indications for Singapore HSA-approved medications.

"TxGNN is the first foundation model designed for clinician-centered drug repurposing, integrating knowledge graphs and deep learning to predict drug efficacy for rare diseases." — Huang et al., Nature Medicine (2023)

Learn More

View Methodology

Data Sources

TxGNN Harvard Zitnik Lab ClinicalTrials.gov NIH Clinical Trials PubMed Biomedical Literature DrugBank Drug Database Singapore HSA Health Sciences Authority

Disclaimer
This report is for academic research purposes only and does not constitute medical advice. Please follow physician instructions for medication use. Any drug repurposing decisions require complete clinical validation and regulatory approval.

Last Review: 2026-03-03 | Reviewer: SgTxGNN Research Team