Extended Data Fig. 1: Data quality and biological replicates from mid-tailbud stage.

a, b, Distribution plot of reads numbers, UMIs, gene numbers, correlation coefficient (Spearman) and saturation level per cell from mid-tailbud (a, midTII_biorep1; b, midTII_biorep2). c, The first two principal components were plotted for cells regressed by UMIs (midTII_biorep1, n = 4,929 cells; midTII_biorep2, n = 4,062 cells). d, The first two principal components were plotted for cells regressed by both UMIs and batches. e, The first two canonical correlation vectors were plotted after alignment by canonical correlation analysis. f–h, Merged (f) and split (g, h) t-SNE clustering for the biological replicates. i, t-SNE plot of canonical-correlation-analysis-aligned samples of biological replicates (n = 8,991 cells). The numbers indicate different clusters. j, The percentage of cells between replicates within the same cluster (clusters shown in i). k, Box plot of the percentage of cells in each cluster (n = 40 clusters) between replicates. The lower, middle and upper hinges correspond to the first and third quartiles (the 25th and 75th percentiles), and the middle hinge corresponds to the median.