Fig. 4

Detailed characterization of ECs in HPSCC. a UMAP plot of ECs colored by cluster, sample type (up), and subgroup markers (down). b Heatmap showing signature DEGs between three distinct EC subsets. c Bubble heatmap showing the ACTA2, COL1A1, COL1A2, and COL3A1 expression among the different EC subsets. Dot size indicates the fraction of expressing cells, colored according to the z-score normalized expression levels. d UMAP plot of ECs colored by cell type. e Pseudotime trajectory analysis of ECs and fibroblast cells. Each dot represents one single cell, colored according to its cluster label. The inlet plot shows each cell with a pseudotime score from dark blue to light blue, indicating an early and terminal state, respectively. A jitter plot showing expression changes in ACTA2, COL1A1, COL1A2, and COL3A1 over pseudotime. f Representative images showing mIHC staining of VWF and α-SMA in HPSCC tumor samples in individual and merged channels. Scale bar = 50 μm. g Differential pathways enriched in tumor and normal ECs according to GSVA score (two-sided unpaired limma-moderated t-test). h Differences in pathway activities scored per cell using GSVA between blood and lymphatic ECs in tumor versus normal tissues. i Heatmap representing t-values for AUC scores of gene regulation based on transcription factor expression in EC subsets, estimated using SCENIC