Fig. 1: Familial structures and microbiome profiles.
From: Variation and transmission of the human gut microbiota across multiple familial generations
a, Family structures in the CGC cohort (n = 102). The colours represent family IDs and the line type indicates current cohabitation status. b, Geographical distribution of the participants across Flanders (Belgium). Family IDs are coloured as in Fig. 1a. c, Cumulative effect sizes of significant covariates on microbiome community variation (dark blue, right bars; stepwise dbRDA on Bray–Curtis dissimilarity) compared to individual effect sizes assuming covariate independence (light blue, left bars; dbRDA on Bray–Curtis dissimilarity) in the CGC (n = 101) and the ≥4-year-old CGC (4+ CGC, n = 91) cohorts (Supplementary Table 2). Age, SBMI, delivery mode, family ID, cohabitation status, medication use, antibiotic use, moisture content (%) and faecal calprotectin (μg g−1) were tested as potential microbiome covariates. d, PCoA of interindividual differences (Bray–Curtis) in relative microbiome profiles of the CGC cohort (n = 101 samples, larger dots) on a background dataset from a cross section of the Flemish population (n = 1,106 samples, small dots). The distribution of samples along the first axis of the PCoA (bottom box plots) separated young children (<4 years old, n = 10) from other individuals (n = 91) and FGFP samples (n = 1,106); Kruskal–Wallis with post-hoc Dunn test, **Padj< 0.01, ***Padj < 0.001. The body of the box plot represents the first and third quartiles of the distribution and the median line. The whiskers extend from the quartiles to the last data point within 1.5× the interquartile range (IQR), with outliers beyond. Bact1: Bacteroides 1; Bact2: Bacteroides 2. e, Increased ARG load (n ARG/microbial load in Bact2 enterotyped samples; n = 101, Kruskal–Wallis test, chi-squared = 26.7, P = 6.8 × 10−6; post-hoc Dunn test, *Padj < 0.05, **Padj < 0.01, ***Padj < 0.001; Supplementary Table 3. The body of the box plot represents the first and third quartiles of the distribution and the median line. The whiskers extend from the quartiles to the last data point within 1.5× the IQR, with outliers beyond. Colour coding as in Fig. 1d. f, Increased prevalence of the Bact2 enterotype in young children (n = 10) compared to other individuals (n = 91) and FGFP samples (n = 1,106); pairwise chi-squared test, Padj < 0.1; Supplementary Table 4. Colour coding as in Fig. 1d. g, Seven bacterial genera displayed significantly higher abundances among members of specific families compared to the rest of the 4+ CGC dataset (n = 91, two-sided Wilcoxon rank-sum test, −log10(P) > 4.56; Supplementary Table 5). The colours correspond to family IDs (as in Fig. 1a), while the circle sizes are proportional to the average abundance of the genus. The closed circles indicate genera with significantly increased abundances in a specific family; the open circles indicate those with decreased abundances in specific families.
