Identification of Cuproptosis-related asthma diagnostic genes by WGCNA analysis and machine learning.

Cuproptosis is the latest novel form of cell death. However, the relationship between asthma and cuproptosis is not fully understood. In this study, we first screened differentially expressed cuproptosis-related genes from the Gene Expression Omnibus (GEO) database and performed immune infiltration analysis. Subsequently, patients with asthma were typed and analysed by Kyoto Encyclopedia of Genes and Genomes (KEGG). Weighted gene co-expression network analysis (WCGNA) was performed to calculate the module-trait correlations, and the hub genes of the intersection were taken to construct machine learning (XGB, SVM, RF, GLM). Finally, we used TGF-β to establish a BEAS-2B asthma model to observe the expression levels of hub genes. Six cuproptosis-related genes were obtained. Immuno-infiltration analysis shows that cuproptosis-related genes are associated with a variety of biological functions. We classified asthma patients into two subtypes based on the expression of cuproptosis-related genes and found significant Gene Ontology (GO) and immune function differences between the different subtypes. WGCNA selected 2 significant modules associated with disease features and typing. Finally, we identified TRIM25, DYSF, NCF4, ABTB1, CXCR1 as asthma biomarkers by taking the intersection of the hub genes of the 2 modules and constructing a 5-genes signature, which nomograph, decision curve analysis (DCA) and calibration curves, receiver operating characteristic curve (ROC) showed high efficiency in diagnosing the probability of survival of asthma patients. Finally, in vitro experiments have shown that DYSF and CXCR1 expression is increased in asthma. Our study provides further directions for studying the molecular mechanism of asthma.