Fig. 5

a Classification Area under the Curves (AUCs) of Random Forest-based classifiers (trained on the marker profiles) for predicting the different disease classes with respect to healthy controls. For each disease state, classification accuracies were obtained after 100 iterations, where in each iteration, the model was trained on 50% of the data and validated/tested on the rest 50%. b PCA plots of the vectors of the ranked feature importance scores for each iterations for three diseases (300 vectors for 100 iterations for each of the three diseases), showing significantly distinct profiles of the feature importance for classification of the three diseases (PERMANOVA p-value < 2.8e−13). c Variable importance scores of the markers identified to be optimal for at least one of the three comparisons (CT v/s T2DM, CT v/s T2DM_CAD and CT v/s CAD). d Fold change of the median abundances of the corresponding markers for each disease state versus the controls