Fig. 3.

Comparison of the protein sequence usage (FM score) between the core and rare background species in healthy children. (A) Heatmap visualisation of the mean differences in FM scores between the core (left) and rare (right) background taxa as defined by the 25th species abundance percentile and based on both reference databases. The metabolic pathway analysis used a homology search by tblastn (bitscore ≥ 200 and a coverage ≥ 75%) in which the DNA sequence of the corresponding reference genome was screened for matching protein reference sequences. Per reference metabolic pathway, a matching score (FM score) between 0 (green) and 100 (black) was assigned based on the percentage of the number of sequences per sub reaction that were detected in the corresponding reference genomes. The heatmap hence visualises the mean FM score obtained for the core species community (left) and the rare species community (right) for each normalisation strategy, separately. All the displayed pathways (n = 108) exhibited a mean group difference in their FM scores of at least 10%. The metabolic pathways with the corresponding MetaCyc superclasses [44] can be obtained from Supplementary Table S4. The statistical difference between the core and rare species biosphere was obtained with a rank-based Mann-Whitney U test (p-value < 0.001). (B) A principal component analysis (PCA) was performed based on the scaled and centred FM scores obtained from all 108 metabolic pathways defined in Fig. 3A. (C) Visualisation of the PCA variables in terms of Metacyc superclasses [44] contributing to the separation of the core and rare species biosphere in the first dimension (Dim1) of the PCA in Fig. 3B. The negative values (orange colour) on the x-axis reflect a positive contribution of the separation towards the negative coordinate space (core species) whereas positive values (blue colour) correspond to a positive contribution towards the positive coordinate space (rare species).