Skip to main content
Fig. 2 | Genome Medicine

Fig. 2

From: Circulating bacterial signature is linked to metabolic disease and shifts with metabolic alleviation after bariatric surgery

Fig. 2

Bacterial blood signature derived metabolic syndrome classification index. Probability of being classified as a subject with metabolic syndrome using random forest classification of phyla and genera abundances (n phyla = 20, n genera = 314, in 47 subjects—36 with and 11 without MetS, missing n = 1, nondiabetic, not unambiguously categorized). A, B Corresponding area under the curve for both genus- and phylum-level random forest models for metabolic syndrome prediction respectively. C, D MetS-IGenus and MetS-IPhylum respectively in subjects with and without metabolic syndrome, boxplots with Tukey-whiskers and mean (◆) are shown. Unpaired samples Wilcoxon signed-rank test is used to compare groups and nominal p-values are shown. E, F Spearman’s rank correlations of MetS-IGenus and MetS-IPhylum at baseline with host variables from all timepoints. Nominal significant values are indicated: *p < 0.05, **p < 0.01, ***p < 0.001

Back to article page