Identification of key anoikis-related genes and immune cell infiltration characteristics in T2DM based on integrating bioinformatic analysis and machine learning

Background: The increasing incidence of type 2 diabetes mellitus (T2DM) is a serious threat to human health and poses a significant global economic burden. Anoikis is a special type of apoptosis. There is growing evidence that anoikis plays a key role in the pathogenesis of diabetes, and can modulate cellular immune responses. This study used bioinformatics techniques to identify diagnostic biomarkers of diabetes. Methods: We downloaded the GSE76894 and GSE76895 from the Gene Expression Omnibus (GEO) database. Anoikis-related genes were obtained from the Gene Set Enrichment Analysis(GSEA) website. We used two machine learning algorithms to screen the key genes and we subsequently constructed a nomogram to provide a diagnostic score for diabetes. Receiver Operating Characteristic (ROC) analysis was performed to assess the diagnostic performance of key genes. We also performed protein–protein interaction (PPI) network and Gene Ontology (GO) enrichment analyses of key genes using clusterProfiler package and GeneMANIA database. In addition, we performed immune infiltration analysis to analyze the differences in immune cells between diabetic patients and healthy individuals, and to analyze the correlation between key genes and immune cells. Finally, we constructed a key gene-miRNA network and key gene-transcription factor(TF) network through online websites. Results: After differential expression analysis, 7 key genes were obtained by us through machine learning algorithms, which are AKT1S1, BMF, ITGB1, PDK4, SNAI2, SRC, and ZNF304. We performed an ROC analysis and the results showed that these 7 genes had good diagnostic performance. In addition, based on these key genes, we analyzed their correlation with immune cells. Finally, we analyzed the regulatory networks of key genes. Conclusion: AKT1S1, BMF, ITGB1, PDK4, SNAI2, SRC, and ZNF304 are candidate key genes for the diagnosis of T2DM. These key genes will provide valuable insights into the pathogenesis and immunotherapy of T2DM.