JMIR Publications

ABSTRACT

Delayed diagnosis and limited access to disease-specific expertise remain major challenges for patients with rare diseases, often due to fragmented research communities and insufficient collaboration among clinicians and researchers. Improving connectivity among rare disease experts is critical for accelerating knowledge exchange, supporting specialist referral, and advancing patient care. To address this need, we developed a network-based approach, called a Rare Disease SME Collaborative Network (RD-SME) that connects Subject Matter Experts (SME) in rare disease field through shared scientific contributions in publications, funded projects, and clinical trials. As a proof-of-concept, we developed a cystic fibrosis (CF) based RD-SME named cfRD-SME to demonstrate the use of RD-SME for supporting CF community. Specifically, cfRD-SME was constructed by integrating co-authorship from CF-related publications, co-investigation from NIH funded CF projects, and co-leadership from CF-focused clinical trials. We clustered CF researchers based on semantic similarity of their research efforts presented in PubMed abstracts, funded project summaries, and clinical trial descriptions, and thereafter large language models (LLM) were applied to summarize their research expertise for each cluster. We performed two case examples to demonstrate the application of using cfRD-SME for supporting cystic fibrosis specialists’ identification and potential research collaboration recommendation. While this case study centers on CF as a proof-of-concept, the RD-SME framework is disease-agnostic and intended for expansion to other rare diseases in subsequent versions.