Fig. 2

Impact of the markers, sequence analysis pipeline tools and filtering method on observed OTU richness after Illumina-amplicon sequencing of the five mock communities. Boxplots represent the variation of OTU numbers as a function of: i) Illumina-amplicon sequencing procedure, based on the rpoB marker or the 16S marker; ii) sequence analyses process, based on the FROGS or DADA2 pipeline; iii) OTU filtering method based on a 0.1% read number threshold or a 1% read number threshold. Each boxplot corresponds to a statistical analysis of the 15 mock community samples (five mock communities, three replicates per mock community); thin purple and red lines correspond to minimum and maximum values, respectively; and the thicker green lines in the boxes correspond to the medians. Red dashed lines correspond to the expected number of OTUs in the mock communities (19 and 17 OTUs for expected_rpoB and expected_16S, respectively)