Fig. 5

Chromatin accessibility dynamics induced by CC16 deficiency and genetic variants. a UMAP visualization of WT and CC16-/- mouse lung nuclei (n = 154,136) across two loading inputs by integrating six animals with three replicates from each group. Nuclei are colored by their predicted cell type. The abbreviation of cell labels was described in Fig. 4 except for aEC (arterial endothelial cells), Endo-like (endothelial-like cells), and PNEC (pulmonary neuroendocrine cells). b The number of differential peaks identified between CC16-/- and WT samples for each cell type. The blue bars indicate the peaks less accessible in the knockout samples and the red bars represent the more accessible peaks. c Aggregated chromatin accessibility surrounding the Scgb1a1 (CC16 gene) locus in club and goblet cells per sample. The aggregated accessibility signal for each sample was normalized by the scaling factor that was computed as the number of cells in the sample multiplied by the mean sequencing depth for the cells in that sample. The WT tracks are labeled in blue and the knockout ones are in red. The genomic regions for the significantly less accessible peaks identified in CC16-/- samples per cell type are highlighted by green shade (five peaks in club cells and two peaks in goblet cells). The associated adjusted p-value is shown above the tracks at their corresponding peak region. Adjusted p-values less than 0.0001 are given four asterisks. The peak annotation for the promoter region of Scgb1a1 is colored red. d Chromosomal distribution of the midpoint of differential peaks identified on chromosomes 8 and 19 with the genomic location and density estimate plotted on the x- and y-axis, respectively. e Chromosomal distribution of SNVs identified on chromosomes 8 and 19 for both WT (blue) and CC16-/- (red) samples. The regions between the dashed lines indicate the SNV hotspots where the knockout samples exhibited a substantially higher number of SNVs than WT samples. The y-axis shows the Phred-scaled quality score generated by BCFtools. f Heatmap showing the Jaccard similarity between the hotspot SNVs identified in CC16-/- lungs and the SNPs derived from 36 different strains on chromosome 8 (lower triangle) and 19 (upper triangle). g “Functional” motifs for which gains or losses of the motif instances are associated with significant changes in chromatin accessibility. The motifs associated with increased chromatin accessibility (“opening”) are shown in red and those associated with decreased chromatin accessibility (“closing”) are colored in green. The y-axis represents the Student’s t-test statistic value. Two motif families (one for transcriptional activators and one for transcriptional repressors) are highlighted on the x-axis. h Cell-type-specific enrichment for the motifs that explain chromatin accessibility changes in SNV hotspots. The bar plot next to the enrichment heatmap shows the total number of differential peaks located in the SNV hotspots for each cell type, which is stratified by the peaks that can be explained by the SNV-driven difference in motif presence (red) and unexplained peaks (blue). The values next to the bars denote the percentage of peaks explained. Only the cell types with more than 10 differential peaks identified in the SNV hotspots are shown