Examination of the tRNA adaptation index as a predictor of protein expression levels

Phenotypic differences between closely-related species may arise from differential expression regimes, rather than different gene complements. Knowledge of cellular protein levels across a species sample would thus be useful for the inference of the genes underlying such phenotypic differences. dos Reis et al [1] recently proposed the tRNA Adaptation Index to score the optimality of a coding sequence with respect to a species' cellular tRNA pools. As a preliminary step towards a multi-species analysis that would utilize this index, we examine in this paper its performance in predicting protein expression levels in the yeast S. cerevisiae and find that it likely predicts maximal potential levels of proteins. We also show that tAI profiles of genes across species carry functional information regarding the interactions between proteins.