Fig. 1

Workflow Schematic representation of the workflow used in the present study. 1) The input data are the tissue/cell metabolic models and the gene expression values. 2) The network is constructed connecting reagent and product metabolites involved in the same reaction, catalyzed by one or more enzyme, whose expression represents the weight of the edges. Multiple edges connecting two nodes are then merged into single edges. 3) The whole network is summarized through clustering and supernodes collapse. 4) The distance among summarized networks is calculated using the \(\mathcal {M}^{2}\) metric, and used to cluster the samples