Figure 5.

Associations between taxonomic and metabolomic features from non-infected and IAV-infected animals. (a), Hallagram representation for metabolites associations with microbial taxa. Metabolites shown in figure 6 were analyzed. Spearman correlation was used as similarity metric, with negative values represented in blue and positive values represented in red. Significant blocks were numbered in descending order of significance. (b and c), correlations between bacterial metabolites (b) and taxa (c) and infection-related variables. a–c, only taxa differentially represented between mock-infected and IAV-infected mice were taken into account. The false discovery rate (FDR) was controlled using the Benjamini-Hochberg method with alpha 0.05 and the expected false negative rate (FNR) for block associations was the default 0.2. (d), network of correlations between metabolites, taxa, and infection parameters. The shape and color of nodes indicate the group of data (metabolites, taxa, or infection parameters). Edges colors indicate the correlation coefficient (blue for negative and red for positive), whereas edge thickness is proportional to significance (thicker lines indicate lower adjusted p-values). Only associations with adjusted p-value ≤0.05 and correlation coefficient ρ ≥ |0.15| were included. Inset, Subnetwork of direct associations with IPA. Arg, arginine; his, histidine; Tyr, tyrosine; Trp, tryptophan; C0, carnitine; C3, propionylcarnitine; IPA, indole-3-propionic acid; C4, butyrylcarnitine; TMAO, trimethylamine N-oxide.