Diabetes is one of the oldest diseases known in human history.1 Type 2 diabetes mellitus (T2DM) was first defined as a component of metabolic syndrome and is characterized by a lack of insulin secretion by pancreatic β cells, tissue insulin resistance, and an inadequate response to insulin secretion.2 With disease progression, insufficient insulin secretion results in a lack of glucose homeostasis, causing hyperglycemia.3 The leading causes of the T2DM epidemic are the overall increase in obesity, sedentary lifestyle, high-calorie diets, and the aging population, which has quadrupled the prevalence of T2DM.4

A healthy human body carries millions of microorganisms which form a system called the human microbiota. The genomes that form the human metagenome represent a diverse array of microorganisms, including bacteria, archaea, and viruses. Bacteria are the most abundant members of the human microbiota, competing with the number of cells in the human body.5 Intestinal microbiota is a key player to maintain our homeostasis through various pathways affecting the intestinal barrier function,6 immunomodulation,7 liver health,8 and many other areas.

Intestinal microbiota dysbiosis is an essential factor in the rapid development of insulin resistance in T2DM, by possibly altering intestinal barrier functions, metabolic pathways, and host signaling, which can be directly or indirectly related to insulin resistance.9 Thousands of metabolites derived from gut microbiota interact with epithelial, hepatic, and cardiac cell receptors and modulate the host physiology. The exact molecular mechanisms have not yet been discovered; however, the proposed mechanisms include the destruction of the intestinal barrier and its association with insulin resistance.10, 11

Data mining is the extraction or the discovery of knowledge from large data sets and is defined as the process of discovering significant new connections, trends, and patterns by analyzing large amounts of data.12 Data mining and machine learning are routinely employed in bioinformatics due to the accessibility of an immense amount of unexplored data.13 Biological data analysis can lead to discovering new knowledge embedded in datasets and expand our knowledge across various fields of biological sciences such as genetics, neuroscience, microbiology, and medicine.14

We describe below the first use of large-scale metagenomics analysis to find the key differences in microbiota composition in healthy and T2DM individuals using data mining and multivariate analysis to identify the possible biomarkers of T2DM.