A comparative analysis of synonymous codon usage patterns in forty completely sequenced bacterial genomes.

The complete genomic sequences of a large number of bacteria which are now publicly available presented the opportunity to investigate the major factors shaping synonymous codon usage variation within and among these species. To do this, we analysed synonymous codon usage in forty completely sequenced bacterial genomes, both inter- and intra-genomically, using correspondence analysis. Within each genome, we investigated the importance of various mutational biases, including compositional bias and strand bias, in shaping differences in codon usage between genes within a particular species. We found that translational selection plays a crucial role in determining synonymous codon usage in almost all the genomes investigated. The overall picture that is emerging from this study is that no one factor but, rather, a combination of factors influence codon usage in a genome to produce a pattern that is unique to that species. Analysis of synonymous codon usage in a combined dataset of all genes from the forty different species also resulted in some interesting findings. We found that much of the variation among species is related to genomic G+C content. We observed a difference in synonymous codon usage between thermophilic bacteria and non-thermophiles, which is not related to G+C content. This variation is primarily due to a differing usage of arginine and isoleucine codons in the two groups