Figure 3: Relationships between the different omic levels.

a, Venn diagram displaying the total number of mOTUs in the metagenomic and metatranscriptomic data sets. b, Comparison of mean relative abundances of mOTUs inferred from metagenomic and metatranscriptomic reads in all samples. The colour shading of the points indicates correlation across the analysed samples (legend below the plot). mOTUs with high and low relative activities are highlighted, in addition to a Ruminococcus sp. (magenta), which, despite high relative abundances, showed an overall low correlation between the metagenomic and metatranscriptomic relative mOTU abundances across samples (Spearman's ρ = 0.17). c, Distribution densities of metagenomic and metatranscriptomic depths of coverage for all predicted ORFs in one sample (M01.2-V1) with the uniquely identified proteins and their relative abundances (area under the ion chromatography curve) visualized by points on a purple-to-green scale (legend on the right). d, Venn diagram displaying the number of predicted ORFs in sample M01.2-V1 detected on the metagenomic, metatranscriptomic and metaproteomic levels; the circle with the dashed line and the number in parentheses represent the proteins that were not uniquely identifiable based on the masses of the peptides, which are represented by protein groups. e, Metatranscriptomic coverage of functions, grouped by frequency of detection in the metaproteome. Numbers on top of the figures indicate the number of functions in each group. f, Venn diagram displaying the total number of functional categories detected in the metagenomic, metatranscriptomic and metaproteomic data sets. g–i, Comparison of functional profiles in the metagenomic and metatranscriptomic data sets (g), metagenomic and metaproteomic data sets (h) and metatranscriptomic and metaproteomic data sets (i) (mean of all samples). The colour shading of the points indicates the correlations of the relative metagenomic and metatranscriptomic levels (g), metagenomic and metaproteomic levels (h) and metatranscriptomic and metaproteomic levels (i) of each functional category across the analysed samples. Highlighted functions: TIGR04056: highest mean proportion of mapping metagenomic and metatranscriptomic reads; K13953, enolase (microbial) and cupin, gliadin, Bet_v_1 (Bet v 1 allergen family; plant proteins): high representation in the metaproteome; K02358 (elongation factor Tu) and K02406 (flagellin): most stably expressed functions in the metatranscriptomes and metaproteomes; highly active clusters enriched with ribosomal genes (Supplementary Section 11).