CircoTax: a novel circular representation of differentially abundant microbial ranks

Abstract One of the most relevant and widely used applications of NGS-based microbial abundance analysis is the discovery of taxa that appear as differentially represented between ecological niches. In the last 10 years several methods have been proposed to accomplish this task. All such analyses invariantly report the differentially represented taxa at the appropriate taxonomic rank as well as a p-value (usually adjusted for multiple testing) and a fold change, representing the confidence on differential call and the amount of variation recorded, usually in log scale, respectively. This work describes CircoTax, an R function implementing a specialized ggplot2 graph to represent a rank-aware collection of differentially abundant taxa in an intuitive and compact radial form. Specifically, the CircoTax builds a circular plot with sectors encoding the taxonomy depth (from kingdom to genus) and, departing from the center, a number of radial bars that reach the appropriate sector and whose color and transparency are proportional to the log fold change intensity and direction. In addition to the manageability in providing custom tables to plot, our implementation also integrates fast and user-friendly ancillary functions which automate both the differential analysis and the display of the results while leaving to the user the choice of further customizations. The visualization of results is both graphically appealing, allowing to accommodate tens of differentially abundant ranks, and biologically informative since the amount of variation, the direction and rank are easily intelligible for each differentially represented taxa at a glance.