1708-scBaseCount: AI-Curated Single-Cell Data Repository

Paper Talk · Oct 25, 2026 · 18m

The paper introduces scBaseCount, currently the largest public single-cell RNA sequencing repository, encompassing over 502 million cells across 27 organisms and 75 tissues. To assemble and maintain this massive database, researchers developed SRAgent, a hierarchical AI workflow that autonomously discovers, extracts metadata, and annotates datasets from the Sequence Read Archive. Furthermore, the companion processing pipeline scRecounter standardizes raw sequencing reads to eliminate technical artifacts and harmonize both exonic and intronic counts. Utilizing this uniformly processed…