使用AI注释一个单细胞数据，效果如何呢？（文末有AI注释交流群）

TL;DR - A hands-on evaluation uses the bioSkills single-cell skill package with Codex to annotate mouse-heart scRNA-seq clusters. Its marker-based annotations largely matched the source study’s labels, suggesting AI-guided workflows can assist routine cell-type annotation.

• bioSkills supports Seurat and Scanpy workflows spanning QC, integration, clustering, annotation, trajectory inference, cell communication, and multi-omics analysis.
• The test used GSE308859, covering mouse cardiac remodeling across Sham and three post-TAC time points.
• The markers-annotation skill identified cell populations from cluster markers such as macrophage, endothelial, cardiomyocyte, smooth-muscle, and platelet genes.
• The article reports that skill-derived annotations were broadly consistent with annotations based on the paper’s marker genes, while reference-atlas-based cell-annotation was not evaluated.