Fig 7. A bar chart of the important metabolites driving discrimination groups in the PCA scores with their respective standard error over three replicates in the ESI- mode.

Important metabolites were identified as the top 10% of the PCA loadings in the filtered dataset. Metabolites “1b 3a 7a Trihydroxy”, “3a 6b 7b Trihydroxy”, and “Dihydroxy” were abbreviated for “1b 3a 7a Trihydroxy 5b cholan 24 oic acid”, “3a 6b 7b Trihydroxy 5b cholan 24 oic acid”, and “3alpha 12beta Dihydroxy 5beta cholan 24 oic acid”, respectively.