CoCoView - A codon conservation viewer via sequence logos

Abstract

Sequence logos are a simple way to display a set of aligned sequences, and they are useful to identify conserved patterns. Since their introduction, several tools have been developed for generating these representations at the single residue level (amino acids or nucleotides). We have developed a tool to build sequence logos of protein-coding sequences at the codon level, allowing more accurate analysis of coding-sequences as they represent synonymous and non-synonymous changes instead of showing only changes that imply on amino acid substitutions. We built CoCoView on top of the Logomaker Python API. It creates codon sequence logos from a multiple sequence alignment of protein-coding sequences. Some properties of the data and the generated logos can be controlled by the end-users, such as data redundancy, plot type and alphabet color. • Split aligned sequences into codon positions; • For each position compute codon frequency and information content; • Use the computed information to plot the graphic.

Keywords: Codons representation; Consensus sequence; Conserved patterns; Information theory.

Graphical abstract

Fig. 1

CoCoView logo based on a multiple sequence alignment of a region of AP2 transcription factor coding sequences from Nicotiana tabacum. (A) Sequence logo generated using WebLogo , representing a per-nucleotide analysis. (B) Sequence logo generated using CoCoView (per-codon analysis). A per-nucleotide analysis could erroneously suggest that some codons are common, which can be ruled out on a per-codon visualization. Exemplified by the codon “GAT”, at the position highlighted in gray on both sequence logos, which can be interpreted as a common codon in the per-nucleotide analysis. However, in the per-codon analysis, this codon does not occur at this position.