Fig. 1

Classifier performance on mock community datasets for 16S rRNA gene sequences (left column) and fungal ITS sequences (right column). a Average F-measure for each taxonomy classification method (averaged across all configurations and all mock community datasets) from class to species level. Error bars = 95% confidence intervals. b Average F-measure for each optimized classifier (averaged across all mock communities) at species level. c Average taxon accuracy rate for each optimized classifier (averaged across all mock communities) at species level. d Average Bray-Curtis distance between the expected mock community composition and its composition as predicted by each optimized classifier (averaged across all mock communities) at species level. Violin plots show median (white point), quartiles (black bars), and kernel density estimation (violin) for each score distribution. Violins with different lower-case letters have significantly different means (paired t test false detection rate-corrected P < 0.05)