Fig. 3: NMDS clustering based on 16 S relative abundance of taxa in the GM, showing major cluster separations according to Dirichlet Multinomial Mixtures (DMM) clustering.

a Mothers and children group into two clusters, almost entirely separating children (cluster 1, 94.5% of cluster members) from mothers (cluster 2, 99.1% of cluster members). b Mother samples (N = 121 biologically independent samples) grouped into two clusters (n1 = 61 and n2 = 61) but showed no significant differences in social disadvantage (SD) or psychosocial stressors (PS; two-sided Mann–Whitney U-tests with FDR adjustment; P = 0.944 for each comparison). The range of the boxes extends from the 25th to the 75th percentiles, the whiskers extend from the minimum to the maximum values, and the black horizontal lines in the boxes indicate the mean values of the data. c Children samples (N = 121 biologically independent samples) also grouped into two clusters (n1 = 61 and n2 = 61). Cluster 1 has significantly higher SD and PS scores (two-sided Mann–Whitney U-tests with FDR adjustment). The range of the boxes extends from the 25th to the 75th percentiles, the whiskers extend from the minimum to the maximum values, and the black horizontal lines in the boxes indicate the mean values of the data. d The proportion of feedings with breast milk is shown on the NMDS clustering.