Fig. 3: Abundance and richness of Nitrospira (top), Nitrosomonas (middle), and Brocadia (bottom) in the metagenome dataset.

a Relative abundances. b The number of species. c The number of genes. d Number of pangenome gene clusters. a, b singleM data; circles show average values of 14 marker genes, error bars indicate 95% confidence intervals. c, d Kaiju data. Note that Kaiju uses NCBI taxonomy, while singleM uses GTDB taxonomy; in the latter the genus Nitrospira is classified as the family Nitrospiraceae. Pearson correlation was estimated for sidestream exposure from 0 to 100%, for the singleM average (a, b) and the Kaiju values (c, d); an asterisk (*) is used to show significant correlations (p < 0.05).