AI RESEARCH
DOGMA: Weaving Structural Information into Data-centric Single-cell Transcriptomics Analysis
arXiv CS.LG
•
ArXi:2602.01839v2 Announce Type: replace Recently, data-centric AI methodology has been a dominant paradigm in single-cell transcriptomics analysis, which treats data representation rather than model complexity as the fundamental bottleneck. In the review of current studies, earlier sequence methods treat cells as independent entities and adapt prevalent ML models to analyze their directly inherited sequence data.