Jun Ding and his team at McGill University have unveiled DOLPHIN, an artificial intelligence tool that takes single-cell transcriptomics to the next level. Unlike conventional methods that collapse RNA changes into a single count per gene, DOLPHIN zooms in on the smaller building blocks of genes—exons and junctions—to reveal crucial variation often hidden from view.
In a study published in Nature Communications, the team showed that DOLPHIN detected over 800 disease markers missed by traditional approaches in pancreatic cancer cells. The tool distinguished aggressive cancers from less severe cases, offering insights that could help doctors match patients to the right treatments earlier and more accurately.
Beyond immediate applications, DOLPHIN lays the foundation for building virtual cell models—digital simulations of cell behavior that could speed drug discovery and reduce reliance on trial-and-error in the clinic.
Read the Article
DOLPHIN advances single-cell transcriptomics beyond gene level by leveraging exon and junction reads.
Song K, Zheng Y, Zhao B, Eidelman DH, Tang J, Ding J. Nat Commun. 2025 Jul 4;16(1):6202.
In the News
New AI tool detects hidden warning signs of disease. Researchers say a closer look inside cells could be used by physicians to detect diseases earlier and better match patients to therapies. by Keila DePape, McGill Newsroom. October 1, 2025.