COUSIN (COdon Usage Similarity INdex): A normalized measure of Codon Usage Preferences

Abstract Codon Usage Preferences (CUPrefs) describe the unequal usage of synonymous codons at the gene, genomic region or genome scale. Numerous indices have been developed to measure the CUPrefs of a sequence. We introduce a normalized index to calculate CUPrefs called COUSIN for COdon Usage Similarity INdex. This index compares the CUPrefs of a query against those of a reference dataset and normalizes the output over a Null Hypothesis of random codon usage. COUSIN results can be easily interpreted, quantitatively and qualitatively. We exemplify the use of COUSIN and highlight its advantages with an analysis on the complete coding sequences of eight divergent genomes, two of them with extreme nucleotide composition. Strikingly, COUSIN captures a hitherto unreported bimodal distribution in CUPrefs in genes in the human and in the chicken genomes. We show that this bimodality can be explained by the global nucleotide composition bias of the chromosome in which the gene resides, and by the precise location within the chromosome. Our results highlight the power of the COUSIN index and uncover unexpected characteristics of the CUPrefs in human and chicken. An eponymous tool written in python3 to calculate COUSIN is available for online or local use.