Cure MFM13
Drug Repurposing Resourses
‍

Drug Repurposing Resources

‍

<
>

Cure MFM13 is committed to open science and collaboration. We are exploring every feasible avenue to treat this debilitating disorder, including drug repurposing.

‍

What is drug repurposing?

Drug repurposing (also known as drug repositioning or drug recycling) is the strategy of investigating medicines already approved by regulatory agencies - such as the FDA in the United States or the EMA in Europe - for new therapeutic indications they were not originally developed for. Because these medicines have already been through extensive clinical testing in humans, their safety profiles, pharmacokinetics, dosing, and manufacturing processes are well established. This means that if a repurposed drug shows promise for a new disease, the path to making it available to patients can be dramatically shorter, cheaper, and less risky than developing an entirely new molecule from scratch.

For ultra-rare diseases like MFM13, drug repurposing holds particular promise. Developing a new drug for a patient population of approximately 60 individuals worldwide faces enormous commercial and logistical hurdles. But testing whether an existing, affordable, and readily available medicine can address the underlying disease biology is something that can be pursued right now. Successful repurposing can also offer patients an actionable treatment option that does not require waiting for a multi-year drug development pipeline.

‍

Datasets and tools

We are currently reviewing publicly available datasets and computational tools for drug repurposing. We encourage researchers working on MFM13 or related conditions to explore these resources as well.

‍

Every Cure - MATRIX Platform

Every Cure is a nonprofit organization founded by Dr. David Fajgenbaum, who discovered repurposed treatments for his own rare disease (Castleman disease). Their MATRIX platform uses AI and biomedical knowledge graphs to systematically score all-by-all treatment likelihood between approved drugs and known diseases - over 66 million drug-disease pairs. The project is funded by ARPA-H ($48.3 million) and is fully open source. Every Cure has identified or advanced 14 repurposed drugs that have become effective treatments for five rare diseases.

MATRIX is supported by MeDIC (Medicines, Diseases, Indications, and Contraindications), a curated foundational dataset for drug repurposing.

‍

NetMedGPT - Network Medicine Foundation Model

NetMedGPT is a transformer-based foundation model for drug repurposing, built on a large-scale biomedical knowledge graph. Using a masked token prediction approach, it learns complex relationships between drugs, diseases, genes, and biological pathways. NetMedGPT can predict drug indications, contraindications, off-label uses, adverse drug reactions, and drug-target interactions, and it generates interpretable disease mechanism subnetworks for each prediction. It outperforms TxGNN and other existing baselines across multiple prediction tasks.

‍

TxGNN - A Foundation Model for Clinician-Centered Drug Repurposing

TxGNN is a graph neural network capable of zero-shot drug repurposing - predicting therapeutic candidates for diseases that have no existing treatments. Trained on a comprehensive biomedical knowledge graph covering 17,080 diseases, TxGNN includes a human-interpretable Explainer module that generates multi-hop mechanistic rationales for each prediction. TxGNN was validated against real-world off-label prescription data from large healthcare systems.

  • Authors: Kexin Huang, Payal Chandak, Qianwen Wang, Shreyas Havaldar, Akhil Vaid, Jure Leskovec, Girish N. Nadkarni, Benjamin S. Glicksberg, Nils Gehlenborg, Marinka Zitnik
  • Paper (Nature Medicine, 2024)
  • Code - GitHub
  • Organization: Harvard Medical School

‍

Contributing

Contributions are welcome. If you are a researcher, clinician, or patient advocate and you know of additional drug repurposing datasets, computational tools, or screening resources that could be relevant to MFM13 or related neuromuscular conditions, please get in touch at ania@curemfm13.org

‍

This page is a living document. We aim to keep it updated as new datasets, tools, and resources become available. Last updated: June 2026.