Extended Data Fig. 3: Transcriptomic analysis of ddx21 mutant and siRNAi knockdown endothelial cells.
(a) Additional RNA Seq analysis for Fig. 4 zebrafish data. PCA and screeplots of all expressed genes (n = 18,549) before (top) and after (bottom) batch correction. Screeplot representing the percentage of variance explained by each principal component for all expressed genes prior to batch correction. X-axis reflects principal component (1 to 8 displayed) and Y-axis reflects the percentage of total variance explained by each component (top left). PCA of components 1 and 2 for all expressed genes and all samples prior to batch correction. Matched sibling and ddx21uq20bh mutant samples share the same shaped point, and sibling and ddx21uq20bh phenotype is depicted by light grey and dark grey colour respectively (top right). Screeplot representing the percentage of variance explained by each principal component for all expressed genes after batch correction. X-axis reflects principal component (1 to 8 displayed) and Y-axis reflects the percentage of total variance explained by each component (bottom left). PCA of components 1 and 2 for all expressed genes and all samples after batch correction. Matched sibling and ddx21uq20bh mutant samples share the same shaped point, and sibling and ddx21uq20bh phenotype is depicted by light grey and dark grey colour respectively (bottom right). (b) Additional RNA Seq analysis for Fig. 4 HUVEC data (as above). PCA and screeplots of all expressed genes (n = 14,588) before (top) and after (bottom) removing aberrant sample HUVEC_siRNA_01_DDX21_01. Samples of the same treatment group (Control, siRNA-construct-01, siRNA-construct-02) share the same shaped point and greyscale colour respectively, except the aberrant sample (HUVEC_siRNA_01_DDX21_01) which is coloured in red. Screeplot representing the percentage of variance explained by each principal component for all expressed genes after removing aberrant sample (HUVEC_siRNA_01_DDX21_01). PCA of components 1 and 2 for all expressed genes and all remaining samples after removing aberrant sample (HUVEC_siRNA_01_DDX21_01). Samples of the same treatment group (Control, siRNA-construct-01, siRNA-construct-02) share the same shaped point and greyscale colour. (c) Reactome pathways significantly enriched in genes upregulated in the ddx21 mutant zebrafish endothelial cells. Y-axis is pathway or GO term, X-axis is number of genes, bars are coloured according to BH (Benjamini-Hoschberg) adjusted p-value. (d) Upset plot indicating Reactome pathways significantly enriched in genes upregulated in ddx21 mutant zebrafish endothelial cells. Bar plot in upper panel: Y-axis represents the number of genes, X-axis the significantly enriched pathways. Points underneath a given bar indicate genes present in that pathway, a line between indicates the same genes present in more than one pathway. (e) Enrichment plot for p53 pathway (zebrafish data). The green curve corresponds to the calculated enrichment score, vertical black lines the position of individual genes in relation to the ranked gene list. Raw and adjusted p values in red. (f) Reactome pathway enrichment bar plot for upregulated genes upon DDX21-siRNA-knockdown. Presented as in (c). (g) Biological process GO term enrichment bar plot (top n = 15 significantly enriched terms) for in DDX21siRNA knockdown HUVECs. Y-axis is BP term, X-axis is number of genes, bars are coloured according to BH (Benjamini-Hoschberg) adjusted p-value. (h) Enrichment plot of p53 pathway (msigdb hallmark gene set) shows log fold change between DDX21-siRNA-knockdown and WT control samples. Presented as in (e). (i) Biological process GO term enrichment upset plot (top n = 15 significantly enriched terms) in DDX21siRNA knockdown HUVECs. Presented as in (d).
