Fig. 2

Cross-organ differences in translation efficiency partially compensate RNA abundance differences and show association with transcript features. a Venn diagram showing the gene expression overlap (i.e. genes detected at both RPF and RNA level) between kidney (yellow, n = 12,423 genes) and liver (green, n = 10,676 genes). Same cutoffs on RPKM (reads per kilobase of transcript per million mapped reads) were used for both organs. b Scatterplot of kidney-to-liver ratio of mRNA abundance versus translation efficiency (TE) for all expressed genes (n = 10,289), averaged over all timepoints. Corresponding density curves are plotted on the margins. Dashed red lines represent the 2.5 and 97.5 percentiles of each variable and the corresponding fold-change is indicated. Linear regression line is depicted in blue (R² = 0.0009, p = 0.0009). While 95% of genes spanned a 114-fold range in mRNA abundance differences across organs, the same number of genes changed less than threefold in TE, underlining that transcript abundance was the main contributor to divergent gene expression. c Inter-organ Spearman correlation for RNA-seq and RPF-seq samples. Each dot represents the correlation coefficient between kidney and liver for a timepoint and replicate sample. Note that RPF-seq samples consistently correlated significantly better than RNA-seq samples (p < 2.2e-16, n = 24, paired t-test of Fisher-transformed correlation coefficients). d Scatterplot of inter-organ RNA vs. RPF correlation coefficients for each sample separately calculated from all (blue, n = 10,289), from single isoform (red, n = 5815), and from multiple isoform (pink, n = 4474) genes. Consistently better RPF correlation was evident in all cases. e Relative TE in liver vs. kidney (data centred and averaged over all timepoints for all expressed genes, n = 10,289) showed an overall strong inter-organ correlation. Differential TE—defined as having false discovery rate (FDR)-corrected p < 0.01 (Wilcoxon signed rank test on TE) and > 1.5 difference in TE across organs—was apparent for ca. 9% of genes (yellow and green show cases where TE is higher in kidney and liver, respectively, n = 960). f Cumulative distribution of Hellinger distances for genes showing differential TE (red, n = 960), or not (grey, n = 9329), as detected in (e). Hellinger distance was used as a quantitative measure for relative transcript isoform diversity across organs, as described in ‘Results’ and ‘Methods’. The analysis shows that divergent TE correlated with larger diversity in transcript isoform expression (D = 0.0702, p = 3.74e-04, two-sample Kolmogorov–Smirnov [KS] test). g Cumulative distribution of the kidney-to-liver TE ratio for genes whose transcript diversity originated exclusively from the 5′ UTR (identical CDS and 3' UTR, light blue, n = 216; these genes show more TE differences across organs) and genes whose transcripts had identical 5′ UTR (and divergent CDS and/or 3′ UTR, purple, n = 314; these genes show less TE differences across organs). The vertical dashed grey line marks the 1.5-fold difference used to define differential TE (as in (e)). These results suggested that tissue specificity in TE was partially achieved by expressing transcript isoforms that differed in their 5′ UTRs (note the significant shift towards smaller TE differences for genes with identical 5' UTRs). See also Additional file 1: Figure S9