About EuTxGNN
EuTxGNN is a drug repurposing prediction system for European Medicines Agency (EMA) approved drugs, powered by the TxGNN framework.
What is Drug Repurposing?
Drug repurposing (also known as drug repositioning) is the process of identifying new therapeutic uses for existing approved drugs. This approach offers several advantages:
- Reduced Development Time: Existing drugs have known safety profiles
- Lower Costs: Bypasses early-stage development
- Higher Success Rates: Known pharmacokinetics and toxicology
The TxGNN Model
TxGNN is a graph neural network model developed for drug repurposing, published in Nature Medicine (2023).
Key Features
- Knowledge Graph Integration: Combines multiple biomedical databases
- Graph Neural Networks: Learns drug-disease relationships from network structure
- Explainable AI: Provides interpretable predictions with supporting evidence
EuTxGNN Pipeline
EMA Data → Drug Normalization → DrugBank Mapping → TxGNN Prediction → FHIR Output
Data Processing
- EMA Medicines Database: Download authorized human medicines
- Article 57 Database: Additional pharmaceutical product data
- Drug Normalization: Standardize drug names to INN format
- DrugBank Mapping: Map to DrugBank identifiers
Prediction
- Knowledge Graph (KG): Network-based association discovery
- Deep Learning (DL): Neural network score prediction
- Evidence Integration: Combine with clinical trial and literature data
Output
- FHIR Resources: Standardized clinical data format
- Web Interface: Browse predictions online
- SMART App: EHR integration capability
Data Sources
| Source | Description | Update Frequency |
|---|---|---|
| EMA Medicines | Centrally authorized medicines | Daily |
| Article 57 | EU pharmaceutical submissions | Periodic |
| TxGNN KG | Biomedical knowledge graph | Static (2023) |
| DrugBank | Drug database | Periodic |
Technical Stack
- Python 3.12+: Core processing
- TxGNN: Prediction model
- FHIR R4: Clinical data standard
- Jekyll: Documentation site
- GitHub Actions: CI/CD
Limitations
- Predictions are computational and require clinical validation
- Limited to drugs with DrugBank mappings
- Based on TxGNN model trained on US data
- Not all EMA drugs have sufficient data for prediction
Contact
For questions or feedback, please open an issue on the project repository.
License
This project is for research purposes only. See the repository for license details.