Fig. 5: Characterization of specific trajectories for fibroblast and pericyte differentiation.

a tSNE plots of 40,315 fibroblasts, colored by cell type (left) and tissue type (top right). The proportions of cell types in normal and tumor tissues are shown on the bottom right. See Fig. 1b legend for the line indications. b Dotplot showing expression status of selected fibroblast markers in each cell cluster. c Diffusion map of fibroblasts. Dot represents single-cell colored by cluster. The two main differentiation branches are indicated with solid arrows. d PAGA analysis of fibroblast subtypes. Line thickness corresponds to the level of connectivity. e Boxplots showing proportion of CAF3 (left) and CAF4 (right) in normal tissues (N = 4), stage I (N = 16) and stage II/III (N = 44) tumors (two-sided Wilcoxon test). Boxplots show the median (central line), the 25–75% interquartile range (IQR) (box limits), the ±1.5 times IQR (Tukey whiskers), and all data points, among which the lowest and the highest points indicate minimal and maximal values, respectively. f tSNE plots of 11,267 endothelial cells, colored by cell type (left) and tissue type (top right). The proportions of cell types in normal and tumor tissues are shown on the bottom right. See Fig. 1b legend for the line indications. g Heatmap showing differences in pathway activities scored per cell by GSVA between different endothelial clusters. h Violin plots showing the expression distribution of selected genes in the endothelial cell clusters. i Dotplot showing selected ligand–receptor interactions. The means of the average expression level of PDGFB in TECs and interacting receptor, PDGFRA and PDGFRB, in fibroblast subtypes are indicated by color. j Spearman’s correlation between the mean expression of PDGFB in TECs and the ratio of pericyte to endothelial cell in all samples. k Spearman’s correlations between the mean expression of PDGFB in TECs and the proportion of CAF2 (left) and CAF4 (right) in all samples, respectively. l Clustered heatmap showing two-sided Spearman correlation coefficients between proportions of immune and nonimmune cell subtypes in tumor samples. P < 0.05 is used as the cutoff value.