Type 2 Diabetes Mellitus Prediction with Gut Microbes Using Machine Learning Through Shotgun Metagenomic Sequencing

Type 2 diabetes mellitus (T2DM) has become a worldwide epidemic as a result of insulin resistance and insufficient insulin production. New research points to dysbiosis in the gut microbiota as a major contributor to type 2 diabetes. Dysbiosis is an imbalance in gut microbes, metabolic activity, or distribution. Gut microbiome dysbiosis alters intestinal barrier functioning and host metabolic and signaling pathways, which contribute to insulin resistance. This study aims to develop machine learning algorithms capable of predicting type 2 diabetes by utilizing features obtained from shotgun metagenomic sequencing of the gut microbiome. The taxonomic and functional characteristics of the gut microbiota are determined by subjecting them to shotgun metagenomic sequencing. We use metrics such as the area under the receiver operating characteristic curve (ROC-AUC), precision, recall, and F1-score to evaluate the models’ effectiveness. In conclusion, the results of this work show that T2DM may be predicted from gut microbial profiles using shotgun metagenomic sequencing in conjunction with machine learning algorithms. SVM with RBF kernel performed the best with 0.8939 ROC-AUC, followed by random forests with 0.8934 ROC-AUC. These findings lay the groundwork for additional investigations into the impact of gut microbiota on the onset of type 2 diabetes and the creation of methods for detecting it without invasive measures. Exploring possible therapeutic approaches targeting the gut microbiota for T2DM control and validating these models in larger and more varied populations require more investigation.