A New UI for AI-Driven Drug Repurposing in Rare Genetic Diseases: Real-World Impact from the NCATS Translator Program

<p xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" class="first" dir="auto" id="d90562e89">The NIH NCATS Translator program now exposes <b>&gt; 200 million semantically harmonized edges</b>, weaving together genomic, pharmacologic, and clinical evidence from <b>over 20 Knowledge Providers</b>—spanning resources such as <b>ClinVar, DrugCentral, DisGeNET, SPOKE, SemMedDB, and real-world EHR aggregates</b>. Leveraging Translator’s reasoning APIs and graph-machine-learning pipelines, our team has systematically surfaced candidate therapeutics for rare monogenic disorders and validated several in patient-specific studies. To democratize this workflow, we built a web-based interface featuring “Smart Queries” that identify drugs predicted to modulate a target gene or treat a disease. This talk will present (i) clinical successes achieved to date, (ii) a live walkthrough of the new UI, and (iii) practical guidance on integrating Translator into rare-disease research pipelines worldwide—accelerating the path from graph insight to patient impact.