Figure 5
From: Mixed-effect Bayesian network reveals personal effects of nutrition
Two subjects were picked from differently reacting clusters and their personally predicted reaction graphs show differences from the general behaviour in Fig. 3 and with each other. The visualization shows 20 most personally prominent effects (\(\hat{\beta }_{ij} + \hat{b}_{ijk}, i=1,\dots \,5, j=1,\dots ,22, k=1,2\)). The red edges denote increasing effect and the blue edges denote a decreasing effect. Subject 1 is reacting generally with a greater increase in insulin concentration than subject 2, but there are even contradicting effects between the subjects. Vitamin D increases insulin concentration for subject 1 only and the effect of lignin is small but opposite between the subjects. The figure is plotted with iGraph package for R language (v 1.2.6, https://igraph.org/r).
