Estimating Gene Expression and Codon-Specific Translational Efficiencies, Mutation Biases, and Selection Coefficients from Genomic Data Alone ‡

Abstract

Extracting biologically meaningful information from the continuing flood of genomic data is a major challenge in the life sciences. Codon usage bias (CUB) is a general feature of most genomes and is thought to reflect the effects of both natural selection for efficient translation and mutation bias. Here we present a mechanistically interpretable, Bayesian model (ribosome overhead costs Stochastic Evolutionary Model of Protein Production Rate [ROC SEMPPR]) to extract meaningful information from patterns of CUB within a genome. ROC SEMPPR is grounded in population genetics and allows us to separate the contributions of mutational biases and natural selection against translational inefficiency on a gene-by-gene and codon-by-codon basis. Until now, the primary disadvantage of similar approaches was the need for genome scale measurements of gene expression. Here, we demonstrate that it is possible to both extract accurate estimates of codon-specific mutation biases and translational efficiencies while simultaneously generating accurate estimates of gene expression, rather than requiring such information. We demonstrate the utility of ROC SEMPPR using the Saccharomyces cerevisiae S288c genome. When we compare our model fits with previous approaches we observe an exceptionally high agreement between estimates of both codon-specific parameters and gene expression levels (ρ>0.99 in all cases). We also observe strong agreement between our parameter estimates and those derived from alternative data sets. For example, our estimates of mutation bias and those from mutational accumulation experiments are highly correlated (ρ=0.95). Our estimates of codon-specific translational inefficiencies and tRNA copy number-based estimates of ribosome pausing time (ρ=0.64), and mRNA and ribosome profiling footprint-based estimates of gene expression (ρ=0.53−0.74) are also highly correlated, thus supporting the hypothesis that selection against translational inefficiency is an important force driving the evolution of CUB. Surprisingly, we find that for particular amino acids, codon usage in highly expressed genes can still be largely driven by mutation bias and that failing to take mutation bias into account can lead to the misidentification of an amino acid’s “optimal” codon. In conclusion, our method demonstrates that an enormous amount of biologically important information is encoded within genome scale patterns of codon usage, accessing this information does not require gene expression measurements, but instead carefully formulated biologically interpretable models.

Selfing in Haploid Plants and Efficacy of Selection: Codon Usage Bias in the Model Moss Physcomitrella patens

“…Therefore, we also used a model-based analysis that is able to model the united effect of mutation bias, natural selection, and their often nonlinear dependence on gene expression level on synonymous codon usage (Gilchrist 2007; Gilchrist et al 2015). Results of this analysis provide further support to the observation that the effect of natural selection on synonymous codon usage is rather weak in P. patens .…”

“…In brief, codon-specific translational inefficiency (η) is a parameter describing how codon usage of a given open reading frame alters the ratio of the expected cost of protein production over the expected benefit of protein synthesis (Gilchrist et al 2015). It is assumed that natural selection favors codon usage that reduces this cost–benefit ratio and selection increases with protein production rates.…”

Analysis of computational codon usage models and their association with translationally slow codons

“…For example, some 24 codon usage models such as %MinMax [9, 10] rely on genome-wide ("ORFeome") CUB. 25 Other models have claimed that codon usage observed in highly expressed genes should 26 be the best indicator of a given codon's translation speed, as highly expressed genes are 27 likely under selective pressure for efficient translation [11,12].…”

“…tAI 134 values are obtained from [21]. ∆η and ∆M values, necessary for calculating codon usage 135 frequencies at varied expression levels per ROC-SEMPPR [11], are from Gilchrist 136 (personal communication). To determine highly expressed codon usage frequencies 137 ("High-Phi", for use in High-Phi %MinMax and High-Phi CAI) per [11], phi was set to 138 5.623.…”

“…Each classifier is trained and tested using 175 5-fold cross validation, and the classes are balanced during training to avoid overfitting. 176 In our analysis: true positives are sequence positions that are predicted as slowly In this work we analyze five computational models related to codon usage (ORFeome 190 %MinMax [9], High-Phi %MinMax [22], High-Phi CAI [22], traditional CAI [12], and High-Phi CAI, based on ROC-SEMPPR high expression ("High-Phi") codon usage 197 estimates [11]. We showed that both models correlate equally as well with empirically 198 measured protein expression in S. cerevisiae [22] as traditional CAI, a model based on 199 CUB in highly expressed genes which is also considered.…”