Skip to main content
. 2021 Sep 23;11:18964. doi: 10.1038/s41598-021-98640-2

Figure 1.

Figure 1

Serum lipid metabolites and bile acids are predictors of the fasting state. (A) Stepwise logistic model parameters predicting the fasting state using 12(13)-EpOME, GCDCA and NO-Gly. (B) Model statistics. (C) Visualization of the correlative environment (generated using hierarchical clustering) of metabolites used for fasting state prediction. Nodes represent branching points in the hierarchical clustering network with metabolites on the fringe named. Metabolite used in the final model are indicated by colors. Directionality of changes in metabolites due to non-fasted state compared to the fasted state are indicated by arrows.