Extended Data Fig. 3: Clustering analysis of two scRNA-seq data across six zebrafish developmental stages and its application for cell type annotation of scATAC-seq data in 10hpf.
a, Uniform manifold approximation and projection (UMAP) visualization of three scRNA-seq data from Farrell, et al.7, Wagner DE, et al.8 and Farnsworth DR, et al.6, colored by samples. Main cell types were labeled in the corresponding region. b, Distribution of maximum prediction scores of cells calculated by the label transfer algorithm in Seurat22,23. c-d, t-SNE visualization of the cells at 10hpf colored by clusters from unsupervised clustering, with two resolutions. e, t-SNE visualization of the cells at 10hpf, colored by gene activity scores of marker genes of yolk syncytial layer (YSL). f, Confusion matrix comparing annotation of scATAC cells using marker genes and labeling of scATAC cells with label transfer. g, Dot plot showing the cell ratio of each cell type between scATAC-seq data (this study) and scRNA-seq data from Wagner DE, et al.8 and Farrell, et al.7 in 10hpf. Linear regression line and 95% confidence interval were shown (Pearson correlation, r = 0.955). Source numerical data are provided as source data.
