Selective Microbial Biomarkers in Type-2 Diabetes with Principal Component Analysis and Receiver-operating Characteristic Curves

Background and aim: Gut microbiota dysbiosis has been associated with metabolic disorders, such as obesity and Type-2diabetes Mellitus. This study evaluated the sensitivity, specificity, and diagnostic accuracy of selective biological markers in T2 diabetes.
Materials and methods: Stool samples were collected from 110 confirmed T2DM and ten non-T2DM subjects, and bacterial DNA extracted. The V4 areas of bacterial 16S rRNA were amplified and sequenced using an Illumina NextSeq 500 platform.
Results: There was a strong correlation between the family Streptococcaceae, Sphingobacteriaceae, Alcaligenaceae, Paraprevotellaceae, and Enterobacteriaceae with T2D. The genus-Faecalibacterium and genus-Roseburia demonstrated a negative correlation with T-2D. The Receiver-operating characteristic (ROC) of the Area Under Curve (AUC) value of gut microbiome was in increasing order with family> Genus > Species > Order> Class.Therefore, we classified the diagnostic accuracy as poor (0.6 < ROC AUC ≤ 0.7), failed (ROC AUC ≤ 0.6), good (0.8 < ROC AUC ≤ 0.9), excellent (0.9 < ROC AUC ≤ 1.0) and fair (0.7 < ROC AUC ≤ 0.8).According to the results, the selected bacterial family/taxa provided fair diagnostic tools followed by genus/taxa, whereas other bacterial genera /taxa failed the diagnostic accuracy.
Conclusion: We could demonstrate the gut microbiome-based classifiers' potential for identifying people suffering from the increased risks for T2D. The findings also revealed that genus-Faecalibacterium, genus-Roseburia, and genus-Phascolarctobacterium were the main discriminants for T2D.