A Possible New Biomarker for T1D? : RNA Splicing in Circulating Blood

Webb-Robertson et al described that alterations in RNA splicing in the circulation differ between individuals with and without new-onset type 1 diabetes (T1D), suggesting that the alternative splicing is potentially valuable for developing diagnostic biomarkers and providing insight into the disease pathogenesis (1). First, they collected whole blood RNA from 2 independent cohorts. One contains new-onset T1D, already developed to diabetes with circulating autoantibodies within 48 hours of diagnosis, and the other consists of the condition-matched controls. They found the specific splicing isoforms, especially retained introns, as those with a significant relationship to T1D patients. Then, they applied machine learning with a Random Forest model to the analysis, showing its high predictive performance in distinguishing between T1D subjects and controls. Finally, they performed gene ontology analysis for the retained intron category to show that the pathway for defense response to viruses is enriched, suggesting a potential underlying etiology of T1D.