Fig. 3

Compound network between bacteria genera and metabolites with the best discriminatory ability between dairy cows that developed metritis (n = 52) and dairy cows that did not develop metritis (n = 52) on the day of metritis diagnosis (3, 7, or 10 days after calving) according to the sparce partial least squares - discriminant analysis (sPLS-DA). Compound network analysis was performed between bacteria genera and metabolites with a correlation coefficient (M) > 0.5 or < -0.5 using the Data Integration Analysis and Biomarker discovery using Latent variable approaches for Omics studies (DIABLO) function of the mixOmics package. Metscape 2 within the CytoScape 3.8 platform was used to edit the figure for easier visualization. Green octagons correspond to metabolites that differed between dairy cows that developed metritis and dairy cows that did not develop metritis, while blue octagons correspond to bacteria genera that differed between dairy cows that developed metritis and dairy cows that did not develop metritis. Large octagons represent metabolites or bacteria genera with greater abundance, while small octagons represent metabolites with lower abundance in the uterine fluid of dairy cows that developed metritis when compared with dairy cows that did not develop metritis. Black lines correspond to positive correlation coefficients while orange lines correspond to negative correlation coefficients