Skip to main content
Fig. 6 | Microbiome

Fig. 6

From: Gut microbiota dysbiosis contributes to the development of hypertension

Fig. 6

A classification to identify pHTN and HTN patients from controls. a, b Random forest models are constructed using explanatory variables of CAGs + species (red curve), CAGs + metabolites (green curve), metabolites (yellow curve), CAGs (blue curve), and species (purple curve). The AUC shows the classification of control versus HTN, or control versus pHTN as the numbers of variables increase. The CAGs + metabolites-based classification is more efficient as indicated by a higher AUC. c ROC of the random forest classifier using CAGs + metabolites based on the 1000 most important variables by ranking the variables by importance. AUC = 0.91 for control versus pHTN (n = 12, red curve), AUC = 0.89 for control versus HTN (n = 12, green curve), and AUC = 0.57 for pHTN versus HTN (n = 13, blue curve). d The top 50 different CAGs distinguish HTN from control based on the random forest model using explanatory variables of CAGs + metabolites. e The top 50 CAGs discriminate between pHTN and control using explanatory variables of CAGs + metabolites. The lengths of bar in the histogram represent mean decrease accuracy, which indicates the importance of the CAG for classification. The color denote the enrichment of CAG in control (blue), in HTN or pHTN (red) according to OR score

Back to article page