Integrative analyses of posttranscriptional regulation in the yeast Saccharomyces cerevisiae using transcriptomic and proteomic data

Correlation between mRNA and protein expression is typically modest due to substantial posttranscriptional regulation. Using large-scale transcriptomic and proteomic data of the yeast Saccharomyces cerevisiae, we quantitatively examined the effects of several posttranscriptional biological properties on the correlation between mRNA and protein expression levels (mRNA-protein correlation) on a genomewide scale. The two classes of properties investigated are (1) stability of mRNA and protein molecules and (2) biological properties related to translational process, such as codon usage and amino acid usage, and experimental data of ribosome density and occupancy. The multiple regression analysis showed that while mRNA half-life and translation initiation efficiency (estimated as mRNA secondary structure in the 5'-UTR) do not appear to have remarkable contributions to the variations in the mRNA-protein correlation, protein half-life descriptor (PHD) is identified as the most important property affecting mRNA-protein correlation (contributing to 16.87% of the total variation in mRNA-protein correlation), suggesting protein degradation significantly affects mRNA-protein correlation. Codon usage and amino acid composition contribute to 8.89% and 7.60% of the total variation, respectively, which is consistent with several previous studies in bacteria (such as Escherichia coli, Haemophilus influenzae, and Desulfovibrio vulgaris), suggesting that mRNA-protein correlation is affected the most by elongation during protein translation. Taken together, all posttranscriptional biological properties contributed to 33.15% of the total variation of mRNA-protein correlation.