Metformin

Evidence Level: L5 Predicted Indications: 50

Quick Overview

Item Value
Drug Name Metformin
DrugBank ID DB00331
Brand Names (EU) Metformin
Evidence Level L5
Predicted Indications 50
Top Prediction Score 99.45%

Approved Indication (EMA)

Synjardy is indicated in adults and children aged 10 years and above for the treatment of type 2 diabetes mellitus as an adjunct to diet and exercise:•          in patients insufficiently controlled on their maximally tolerated dose of metformin alone•          in combination with other medicinal products for the treatment of diabetes, in patients insufficiently controlled with metformin and these medicinal


Predicted New Indications

TxGNN model predictions for potential drug repurposing:

Rank Indication Score Source
1 focal stiff limb syndrome 99.45% DL
2 classic stiff person syndrome 99.45% DL
3 diabetes mellitus (disease) 99.43% DL
4 opsismodysplasia 99.40% DL
5 thiamine-responsive dysfunction syndrome 99.40% DL
6 drug-induced localized lipodystrophy 99.06% DL
7 centrifugal lipodystrophy 98.99% DL
8 pressure-induced localized lipoatrophy 98.96% DL
9 pancreatic agenesis 98.91% DL
10 idiopathic localized lipodystrophy 98.90% DL
11 homozygous familial hypercholesterolemia 92.30% DL
12 autoimmune oophoritis 83.94% DL
13 hypervitaminosis 82.40% DL
14 type 1 diabetes mellitus 79.77% DL
15 pulmonary atresia-intact ventricular septum syndrome 77.36% DL
16 pulmonic stenosis (disease) 76.27% DL
17 genetic vascular tumor 73.85% DL
18 arterial duct anomaly 73.33% DL
19 proximal 16p11.2 microdeletion syndrome 73.09% DL
20 subaortic stenosis, membranous 71.88% DL

Showing top 20 of 50 predictions.


About TxGNN Predictions

Prediction Sources

Source Description
KG Knowledge Graph - Network topology-based associations
DL Deep Learning - Neural network score prediction

Evidence Levels

Level Definition
L1 Multiple Phase 3 RCTs / Systematic Reviews
L2 Single RCT or multiple Phase 2 trials
L3 Observational studies / Large case series
L4 Preclinical / Mechanistic / Case reports
L5 AI prediction only (current)

Clinical Validation Needed

Research Use Only: These predictions are computational hypotheses that require clinical validation. They should NOT be used for clinical decision-making.

Next Steps for Validation

  1. Literature Review: Search PubMed for existing evidence
  2. Clinical Trial Search: Check ClinicalTrials.gov for ongoing studies
  3. Mechanistic Analysis: Evaluate biological plausibility
  4. Preclinical Studies: Conduct in vitro/in vivo validation
  5. Clinical Trials: Design and conduct human studies

Data Access


Citation

If using this data, please cite:

@article{huang2023txgnn,
  title={A foundation model for clinician-centered drug repurposing},
  author={Huang, Kexin and others},
  journal={Nature Medicine},
  year={2023},
  doi={10.1038/s41591-023-02233-x}
}

Disclaimer: This report is for research purposes only and does not constitute medical advice. Drug repurposing predictions require rigorous clinical validation before any therapeutic application.

Copyright © 2026 EuTxGNN Project. For research purposes only. Not medical advice.