Fig. 3
TDL inputs and predictions. a Volcano plot of the P-values and the fold changes of the ileum transcriptomics data. mRNA species highlighted in green are those with P-values <0.05 after FDR correction and with fold change >1.5, i.e., higher in GF. mRNA species highlighted in red are those with P-values <0.05 after FDR correction and with fold change <−1.5, i.e., higher in M8 mice. b Legacy knowledge about impact of E. coli on alterations in lipid metabolism included impacts mediated by inflammation and alterations in dietary lipids. c Coupling differential gene expression data from ileum transcriptomics and organism level databases (lipid specific e.g. LIPID MAPS and non-lipid specific, e.g., Reactome) allowed us to predict using different genes potential alterations in lipid metabolism. d Coupling differentially expressed genes in the ileum to genome-scale metabolic models (GEMs), we could predict potential alterations in terms of biochemical transformations (reactions) and participating metabolites (including lipids). Compiling the predictions from different branches of TDL, we formulate the predicted search space of altered lipids for further measurement and analysis. e Comparison of differentially expressed genes implicated by different aspects of TDL. GF germ-free, M8 GF mice monoinoculated with M8 strain of E. coli, GEMs genome-scale metabolic models
