Extended Data Fig. 1: Cell clustering analysis of human fetal cerebellar cells and NSC and TCP comparation.

a, Comparison of the number of genes/cell (left) and number of read counts/cell (right) between the sc-RNA-seq of human fetal cerebellar cells and published snRNA-seq dataset⁵. Unpaired two-tailed Student’s t test. b, UMAP clustering of human fetal cerebellum single cell samples. Colors indicate distinct populations from 95,542 single cells across eight developmental stages. c, t-SNE maps of cell populations at indicated developmental stages. One tissue at each stage was used for the scRNA-seq analysis. d, STREAM plot visualization of the cell density along the indicated cell lineage trajectories. e, Upper, a zoom-in view of lineage trajectories. Lower, volcano plot showing the differentially expressed genes between NSC and TCP. The genes were colored based on fold change differences (>1.5 folds) and FDR-p value (p < 0.05). The statistics was obtained by the two-sided Wilcoxon test as implemented by Seurat. f, Upper left, UMAP plots of NSC and TCP cell populations. Upper right and lower left panels indicate expression of G1/S and G2/M marker genes in NSC and TCP populations, respectively. Lower right, immunostaining for Ki67/HNRNPH1 in the RL transitional zone at PCW 12. Arrows: Ki67 and HNRNPH1 co-stained cells. Scale bar: 50 μm. g, Correlation analysis of the identified progenitor populations in the human fetal cerebellum.