ARPA-H Awards $35 Million to UNC, Emory to Build Rare-Disease AI Data Resource

Rare Daily Staff

The Advanced Research Projects Agency for Health’s Rare Disease AI/ML for Precision Integrated Diagnostics program, known as RAPID, awarded up to $35 million to the University of North Carolina School of Medicine and Emory University to build what the institutions describe as the world’s largest data resource for artificial intelligence research in rare diseases.

The four-and-a-half-year initiative aims to bring together clinical, genetic and other health information that could help doctors diagnose rare conditions sooner, guide patient care and speed the search for new treatments.

UNC and Emory will lead efforts to acquire data from patient registries and real-world sources, while other parts of the program will focus on gathering information directly from patients and building a research platform for the broader rare-disease community.

Melissa Haendel, a genetics professor at the UNC School of Medicine and the project’s lead scientist, said the goal is to develop AI models using unusually detailed data from approximately 2,700 patients. Those models could then support diagnoses in settings that lack specialized rare-disease expertise.

“If we can build models based on the rich data, we can build clinical decision support tools that either diagnose patients in less sophisticated settings or help move those patients along,” Haendel said in the announcement.

Even as genomic testing has become more common, many rare diseases remain difficult to identify. Relevant information may be scattered across electronic health records, insurance claims, imaging files, genetic tests, patient surveys and specialty clinics. Because individual disorders affect relatively few people, no single health system or researcher may have enough data to recognize meaningful patterns.

The UNC-Emory effort plans to combine information from health records, insurance claims, medical imaging, video technologies and patient surveys. Researchers said names and other direct identifiers would be removed before information enters the RAPID data resource.

Access would be tiered according to data sensitivity. Some datasets could be publicly available, while researchers seeking access to more sensitive information would need data-use agreements and additional safeguards. Qualified physicians and researchers worldwide would be able to apply for access.

“Much of the information about rare disease patients isn’t centrally available because each condition affects relatively few patients,” said Richard Moffitt, an associate professor of hematology and medical oncology at Emory University School of Medicine and the project’s co-lead.

Moffitt also cited limited access to specialty care, insurance barriers and patients’ distance from rare-disease experts as factors that can hinder efforts to collect the high-quality data needed to understand disease patterns and outcomes.

The resource could also help address a persistent challenge in rare-disease research: finding eligible participants for clinical trials. Haendel said industry sponsors regularly approach UNC about trials, but identifying patients who meet narrow enrollment criteria can be difficult.

By securely linking larger and more diverse data sources, the team hopes to develop algorithms that can identify potential trial participants more efficiently, strengthen studies and broaden access to specialized care.

The project includes academic institutions, patient-advocacy groups, health-data companies and industry partners. Collaborators include Global Genes, publisher of Rare Daily; the National Organization for Rare Disorders; Johns Hopkins University; the University of California, San Francisco; the University of Iowa; Datavant; Truveta; and Queen Mary University of London. OpenAI, Anthropic, Amazon Web Services and Google are also expected to provide program support.

Photo: Melissa Haendel, a genetics professor at the UNC School of Medicine and the project’s lead scientist