Abstract : Codon usage is biased between lowly and highly expressed genes in a genome-specific manner. This universal bias has been well assessed in some unicellular species, but remains problematic to assess in more complex species. We propose a new method to compute codon usage bias based on genome wide translational data. A new technique based on sequencing of ribosome protected mRNA fragments (Ribo-seq) allowed us to rank genes and compute codon usage bias with high precision for a great variety of species, including mammals. Genes ranking using Ribo-Seq data confirms the influence of the tRNA pool on codon usage bias and shows a decreasing bias in multicellular species. Ribo-Seq analysis also makes possible to detect preferred codons without information on genes function.
https://hal-lirmm.ccsd.cnrs.fr/lirmm-01610054
Contributeur : Eric Rivals
<>
Soumis le : mercredi 4 octobre 2017 - 12:22:35
Dernière modification le : lundi 26 février 2018 - 13:56:02
