Fig. 4: Remodeling of translation by CR.

a Venn diagram showing the overlap of genes that yield 3′UTR-derived reads in all samples from individual conditions. b Principal component analysis of 3′UTR-to-CDS read ratio in all samples of all four conditions. c Violin plots showing the distribution of average 3′UTR-to-CDS read ratio for individual mRNAs across all samples of a given condition. P-values from the Mann–Whitney U-test indicate the significance of the differences in mean values between conditions. d Distribution of changes in the 3′UTR-to-CDS ratio in the three conditions (AGE, CR, RM defined in the text) for the 61 genes with a significant change (p-value < 0.05 from Welch’s t-test) in at least one condition. Scale of variation is shown in the center, with values capped at ±2 in log base 2 fold-change (4-fold increase or decrease in 3′UTR-to-CDS ratio). e IGV³⁹ snapshots of the Hspb7 and Pvalb genes, showing the coverage in the different types of samples. f Functional annotation of genes that show a significant change in 3′UTR-to-CDS ratio in the CR condition, among the genes for which data were available in all samples and all conditions (53 of the 72 genes from the center of panel a). g Meta-gene plot showing the normalized gene coverage around start and stop codons. For each gene, the RPF counts mapping to each nucleotide within a window of ±200 nucleotides of the start and stop codons were normalized, and then accumulated across genes. These profiles are shown along with error bars, calculated by standard deviation across the samples from the same cohort. Highlighted are the region before the stop codon (orange dashed line) and the 3′UTR region immediately downstream of the stop codon (orange full line), where the largest AGE and CR-induced coverage changes are observed.