Fig 1.
Spatial heterogeneity of microbial composition along the gastrointestinal (GI) tract.
Ten samples were collected along the GI tract per individual (A). The averages of relative abundance of bacterial families (B), the microbial diversity measures (C), and the first principle component value from predicted gene function categories (D) across the GI tract are shown. The error bars are standard deviations. Asterisks denote unclassified families within the listed order.
Fig 2.
The relative proportions of the most abundant metabolism related KEGG pathways (level 3) predicted by PICRUSt between upper and lower GI tract.
The error bars are standard deviations. The star indicates Bonferroni corrected P < 0.0033 using Wilcoxson rank sum test.
Fig 3.
PCoA plots of microbial community membership (unweighted UniFrac distance).
Each dot represents a microbial community from one gut segment in one individual. The first principle component (PC1) mostly accounts for differences between the upper and lower GI tract (A). Within the upper GI tract, microbial communities were grouped by host individual to some degree (B). Within the lower GI tract, microbial communities were strongly grouped by host individual (C).
Table 1.
Variables explaining gut microbial communities based on ADONIS.
Fig 4.
Microbial community membership is strongly associated with host individual in the lower GI tract.
Tree is based on UPGMA clustering of unweighted UniFrac distance. Different colors show different gut segments (see Fig 3A). Larger node sizes indicate stronger jackknife support. The brackets show the clustering by individuals in the lower GI tract with a jackknife support of 1.0.