Loss of the chromatin modifier Kdm2aa causes BrafV600E-independent spontaneous melanoma in zebrafish
Fig 3
RNA-seq profiling of individual kdm2aa-deficient embryos and their siblings.
(A) Summary of experimental design. TruSeq stranded mRNA sequencing libraries were prepared from four individual embryos from each genotype at both 5 d.p.f. and 12 d.p.f., totalling 48 libraries. Paired end sequencing with a read length of 75 bp was performed on four lanes of Illumina HiSeq 2500 machines. Sequence was aligned to the GRCz10 reference genome with TopHat and read counts were obtained with HTseq-count. Quality control was performed using QoRTs and outliers removed. 32,261 genes were detected. DESeq2 was used to determine DE genes in each individual clutch, DE genes at each stage by combining the two clutches for each stage whilst controlling for clutch in the DESeq2 model, and also in a combined analysis combining all 4 clutches whilst controlling for stage and clutch. (B) Venn diagram showing the overlap in DE genes between the two 5 d.p.f. experiments and the combined 5 d.p.f. analysis. The increase in power due to the increased sample size enabled an additional 1106 DE genes to be detected. (C) Venn diagram displaying the overlap in DE genes between the two 12 d.p.f. experiments and the combined 12 d.p.f. analysis. An additional 1102 DE genes were detected in the combined analysis. (D) Volcano plot of all detected genes in the combined RNA-seq analysis with the 3752 DE genes shown in red and kdm2aa circled. (E) Box plot of normalised counts for the chromatin modifier nsd1b, with adjusted p-values for individual experiments, stage-specific and combined analysis as indicated by horizontal bars. Data for heterozygous and wild-type siblings are combined. In the figure legend ‘s’ denotes siblings and ‘m’ homozygous mutants. (F) Box plot of normalised counts for dnmt3bb.2, a de novo DNA methyl transferase, with adjusted p-values for individual experiments, stage-specific and combined analysis as indicated by horizontal bars. Data for heterozygous and wild-type siblings are combined. In the figure legend ‘s’ denotes siblings and ‘m’ homozygous mutants. (G) Table of enriched Gene Ontology (GO) terms of the biological process (BP) GO domain which overlap in both 5 d.p.f. experiments and also the 5 d.p.f. combined analysis. P-values shown are for the combined analysis. (H) Table of enriched GO terms that overlap in both individual 12 d.p.f. experiments and the 12 d.p.f. combined analysis, with p-values for the combined analysis shown.