Skip to main content
Fig. 3 | Genome Biology

Fig. 3

From: Translational contributions to tissue specificity in rhythmic and constitutive gene expression

Fig. 3

Rhythmicity analyses across organs reveals phase modulation by translation in kidney. a Venn diagram showing rhythmic genes in kidney. Of the 12,423 expressed genes, 1338 showed 24-h oscillations of > 1.5-fold amplitude in mRNA abundance (RNA-seq, 10.7%) and 977 in footprint abundance (RPF-seq, 7.9%). A total of 542 genes (4.3%) were identified as rhythmic at both levels. b Cumulative distribution of phase differences (RPF peak – RNA peak, in hours) for genes rhythmic at both RNA-seq and RPF-seq in liver (green, n = 1178) and kidney (yellow, n = 542). The two distributions were significantly different (p < 1e-04, permutation test) and reflected that maximal footprint abundance frequently preceded mRNA abundance peaks in kidney (note that the two distributions differed mostly in their negative tail). c Four-way Venn diagram of rhythmicity sets for genes expressed in both tissues (n = 10,289). Of all genes, 364 and 238 were detected as rhythmic in both organs at the RNA-seq and RPF-seq levels, respectively, and 178 genes were detected as rhythmic throughout (i.e. RNA-seq and RPF-seq, in kidney and liver). d Cumulative phase difference distribution in liver (green) and kidney (yellow) for the 178 common rhythmic genes. As in (b), the distributions were significantly different (p = 0.007, permutation test) and corroborated that even when comparing the same set of genes, footprint peaks frequently preceded mRNA abundance maxima in kidney. e Cross-correlation in kidney (yellow) and liver (green) of time-resolved RPF-seq profiles relative to the RNA-seq profiles of the n = 178 common rhythmic genes. The analysis showed that profile correlations for negative lags (i.e. RPF peaking before RNA) were significantly higher in kidney than liver (* indicate p < 0.05, Wilcoxon signed rank test). Boxplots represent the interquartile range and whiskers extend to the minimum and maximum expression within 1.5 times the interquartile range. f Examples for genes with maxima in RPF (blue) preceding those in RNA (orange) by several hours in kidney (top) but not, or less so, in liver (bottom). Arrowheads indicate the peaks in footprint and mRNA abundance as estimated from the rhythmic fits. g Cross-correlation analyses of RPF-seq relative to RNA-seq profiles (kidney in yellow, liver in green) for the genes in (f). Maximal correlations of the profiles in kidney were found to be shifted to the left (more negative RPF-to-RNA lags) as compared with liver. For liver, there was no case with a maximal correlation value in negative RPF-to-RNA lags

Back to article page