Data Downloads
Open data for research use under CC BY 4.0 license.
Quick Downloads
Summary Data
| File | Format | Description |
|---|---|---|
| Drug Mapping | CSV | DrugBank mapping for 638 drugs |
| Predictions | CSV | All 32,368 predictions |
| Indication Mapping | CSV | Disease mapping results |
Complete Dataset
Data Schema
repurposing_candidates.csv
| Column | Description |
|---|---|
license_id |
EMA product number |
brand_name |
Commercial name |
ingredient |
Active substance (INN) |
drugbank_id |
DrugBank identifier |
potential_indication |
Predicted indication |
source |
Prediction source (KG or DL) |
score |
Prediction confidence (0-1) |
drug_mapping.csv
| Column | Description |
|---|---|
ingredient |
Active substance name |
drugbank_id |
Mapped DrugBank ID |
drugbank_name |
DrugBank drug name |
match_type |
Mapping method used |
FHIR Resources
FHIR R4 resources are available at:
| Resource | Endpoint | Count |
|---|---|---|
| CapabilityStatement | /fhir/metadata | 1 |
| MedicationKnowledge | /fhir/MedicationKnowledge/ | 733 |
| ClinicalUseDefinition | /fhir/ClinicalUseDefinition/ | 32,368 |
External Data Sources
These datasets were used to build EuTxGNN:
| Source | URL |
|---|---|
| TxGNN Knowledge Graph | Harvard Dataverse |
| EMA Medicines Database | EMA Website |
| DrugBank | DrugBank |
Citation
When 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}
}
License
All EuTxGNN generated data is released under CC BY 4.0.
Original TxGNN data is subject to its own license terms.