Fig. 2: Lung endothelium exhibits an immediate bimodal response pattern.

a–c, Differential gene expression analysis of lung EC pseudobulks comparing day 0 versus day 1.5 and day 1.5 versus day 3.5 using DESeq2. a, Heatmap of highly differentially expressed genes across the experimental timeline, with log₂ fold change (FC) >3.5 and P <0.01, as computed in DESeq2 using the Wald test. Scaled expression is shown. b,c, Volcano plots of differentially expressed genes for comparison of day 1.5 versus day 0 (b; full set of DEGs shown in Supplementary Table 3) and for comparison of day 3.5 versus day 1.5 of total lung EC pseudobulks (c; full set of DEGs shown in Supplementary Table 4). FC and P values were computed in DESeq2 using the Wald test. Dotted lines represent thesholds of significance. Horizontal dotted lines indicate fold changes > |0.5|, vertical dotted lines indicate P values < 0.01. d, Differentially expressed genes were calculated using gCap and aCap pseudobulks comparing day 0 with days 1.5 and 3.5, as well as day 1.5 with day 3.5. Numbers of significant DEGs for each comparison and EC type are shown. DEGs were computed using the Wilcoxon rank-sum test. Genes were considered significant for log₂ FC > 0.5 and P < 0.01. e, UMAP showing SNN-based clustering of total lung ECs resolved four cell clusters, which were annotated using the same marker genes for the classification shown in Extended Data Fig. 3. The activated gCap cluster was annotated based on upregulated genes compared with the gCap cluster (full set of DEGs shown in Supplementary Table 5). f, UMAP of total lung ECs split by time point and colored by cluster identity (left); barplot showing cell number per cluster and time point (right). The dataset contains 2,479 cells; n = 3–4 mice per time point. g, UMAP of filtered capillary lung ECs, colored by classified cell type. The dataset contains 2,293 cells. h,i, visualization of gene expression as gene scores in UMAP for the biosynthesis gene set (h) and angiokine gene set (i), split by time point.