Fig. 6: Predicting depression status of IBS patients from multi-omics features.

a Receiver operating characteristic curve (ROC) of the RF model using gut microbiota profiles to classify IBS patients from healthy controls in discovery and validation cohorts. b ROC of the RF model using the fecal metabolomic profiles to classify IBS patients from healthy controls in discovery and validation cohorts. c ROC of the RF model using serum metabolome data to classified IBS patients from healthy controls in discovery and validation cohorts. dāf ROC of the RF model using gut microbiota, fecal metabolomic, and serum metabolomic data, respectively, to train and predict depression status of IBS patients. AUC value is presented in the format of 95% confident interval. For serum and fecal metabolomic data, only the possibility distributions derived from untargeted positive ion mode are presented.