Help & Documentation

Guides, methodology, and project information

Quick Links

Resource	Description
Methodology	How predictions are generated
User Guide	How to use this platform
About	Project background and team
Privacy Policy	Data handling and privacy

Frequently Asked Questions

What is SgTxGNN?

SgTxGNN is a drug repurposing prediction platform for Singapore HSA-approved medications, based on Harvard's TxGNN model published in Nature Medicine.

How accurate are the predictions?

Predictions should be viewed as research hypotheses. While TxGNN is validated on known drug-disease relationships, individual predictions require clinical validation before any therapeutic application.

What do the evidence levels mean?

Level	Meaning
L1	Multiple Phase 3 RCTs support this use
L2	Single RCT or Phase 2 evidence
L3	Observational study evidence
L4	Preclinical or mechanistic evidence
L5	AI prediction only, no clinical evidence

What does KG+DL mean?

Predictions marked "KG+DL" are validated by both the Knowledge Graph method and the Deep Learning method, indicating higher confidence.

Can I use this data for research?

Yes, SgTxGNN data is available for academic research with proper citation. See Downloads for data files.

How do I report an issue?

Please report issues via GitHub Issues.

Getting Started

For Researchers

Browse Drug Reports for available drugs
Review Methodology to understand predictions
Download data from Downloads
Cite appropriately in publications

For Clinicians

Search for drugs of interest
Review evidence levels and sources
Check Safety Data for interactions
Remember: predictions are for research only

For Developers

Review FHIR API documentation
Test with SMART App integration
Clone from GitHub

Contact

For questions or feedback:

GitHub Issues: Report a problem
Discussions: Ask a question

Disclaimer
SgTxGNN is for research purposes only and does not provide medical advice. Always consult healthcare professionals for treatment decisions.