Fig. 3

Identification of genes regulated at the transcriptional and translational levels during stress. a Workflow describing differential gene expression (DGE) and translational efficiency (TE) analyses using Ribo-Seq and RNA-Seq reads. In each experiment we subsampled the original table of counts as to have the same total number of reads in each Ribo-Seq and RNA-Seq sample considered. This ensured the results would not be biased by lack of statistical power in the samples with less coverage. The data was used to define regulatory classes for different sets of genes. b Correlation between replicates and between RNA-Seq and Ribo-Seq samples. Two representative examples are shown, data is counts per million (CPM). c Definition of regulatory classes after DGE analyses. Transcriptional change: Genes that showed significant up-regulation or down-regulation using both RNA-Seq and Ribo-Seq data. Translational change: Genes that showed significant up-regulation or down-regulation only with Ribo-Seq data. Post-transcriptional buffering: Genes that showed significant up-regulation or down-regulation only with RNA-Seq data. The axes represent logFC between stress and normal conditions. d Fraction of genes that showed translational or transcriptional changes. DGE was performed with the lima voom software and genes classified in the classes indicated in C. See Table S3 for more details on the number of genes and classes defined. e Significant positive correlation in ribosome density changes in the 5’UTR and the CDS for stress vs normal conditions. Data shown is for the complete set of mRNAs. log₂FC (Fold Change) values based on the number of mapped Ribo-Seq reads, taking the average between replicates. f Same as E but for genes up-regulated at the level of translation. There is no positive correlation in this case