Abstract
Meniere disease (MD) is a common inner ear disorder closely related to immune abnormalities, but research on the characteristic genes between MD and immune responses is still insufficient. We employ bioinformatics and machine learning to predict potential biomarkers and characteristic immune cells associated with MD, investigating the Mendelian randomization causation between immune cells and MD, providing new insight for the early diagnosis, prevention, and treatment of MD. We obtained relevant data on MD from the GEO database using R, conducted differential gene analysis, and performed weighted gene co-expression network analysis (WGCNA) to identify genes associated with MD. Moreover, by integrating the selection of core genes from the PPI with machine learning techniques, we predicted potential biomarkers for MD. Simultaneously, conducted immune infiltration analysis of the core genes and identified key immune cell types. Finally, employed Mendelian randomization to comprehensively evaluate the causal relationship between immune cells and MD. Through differential gene analysis and WGCNA, we identified 550 genes associated with MD, with enrichment analysis predominantly focused on pertinent immune responses and related diseases. The protein-protein interaction (PPI) screening and machine learning techniques, we predicted 2 potential biomarkers for MD: CD5 and AJUBA, 3 core immune cell types associated with MD: T cells CD4 memory resting, T cells gamma delta and Dendritic cells activated. Mendelian randomization analysis revealed a causal relationship between 26 types of immune cells and MD. There is a causal relationship between immune cells and MD. CD5 and AJUBA are potential biomarkers of MD, while T cells CD4 memory resting, T cells gamma delta and Dendritic cells activated are core immune cells of MD. These potential biomarkers and core immune cells offer new insights for the early diagnosis, prevention, and treatment of MD.