Screening and identification of key serum biomarkers between PM2.5 and-induced asthma onset.

Objective: The mechanism by which air pollution causes the onset of asthma is complex, and its key targets have not yet been fully identified. In this study, we identified the factors that mediate the relationship between air pollution and asthma. Methods: We screened overlapping genes related to asthma from the Gene Expression Omnibus (GEO) database by integrating differentially expressed genes (DEGs) and weighted gene co-expression network analysis (WGCNA). To further identify hub genes, we used three machine learning methods: least absolute shrinkage and selection operator (LASSO) regression, support vector machine-recursive feature elimination (SVM-RFE), and random forest (RF). We subsequently analyzed the mediating role of these hub genes in the association between PM2.5 and onset of asthma in a study population consisting of 160 participants with asthma and 160 participants without asthma. The analyses included Pearson correlation analysis, logistic regression analysis, and mediation analysis. Results: A total of 715 DEGs were identified, and the results of the WGCNA revealed a significant asthma-associated co-expression grey module containing 118 genes. Among these genes, five hub genes (CEBPE, HDC, IRAK3, PRR4, and SOD2) were selected using three different machine learning methods. These genes were confirmed as independent predictors of asthma through multivariate logistic regression analysis and were significantly correlated with PM2.5 levels. Mediation analysis demonstrated that these genes play a mediating role between PM2.5 and asthma onset. Conclusion: This study provided evidence that CEBPE, HDC, IRAK3, PRR4, and SOD2 mediate the connection between PM2.5 and the onset of asthma.