Screening Oxidative Stress-Related Biomarkers For Periodontitis Using Machine Learning

The development of periodontitis involves oxidative stress, but its exact mechanism has not been fully elucidated. This study aimed to reveal the role of oxidative stress in periodontitis and to search for potential diagnostic markers and therapeutic targets. Transcriptomic data were obtained from the Gene Expression Omnibus (GEO) database for gingival tissues from periodontitis and controls. Through weighted gene co-expression network analysis (WGCNA), differential expression analysis, gene set enrichment analysis (GSEA), we systematically identified oxidative stress-associated key genes in periodontitis. These candidate genes were first screened by machine learning algorithms and then validated through independent external datasets. Differential expression profiling analysis identified 139 significantly differentially expressed genes (DEGs) in periodontitis samples. WGCNA revealed the 154 genes had the strongest link to oxidative stress. Integration of DEGs and co-expressed module genes via Venn diagram screening yielded 13 oxidative stress-related signature genes. Subsequently, feature selection using three machine learning algorithms ultimately identified CXCL6, COL4A2, and CYR61 as core hub genes regulating the oxidative stress microenvironment in periodontitis, with their diagnostic efficacy validated by ROC curve analysis. Through systematic analysis, this study reveals the important role of oxidative stress in the development of periodontitis and identifies CXCL6, COL4A2, and CYR61 as novel therapeutic targets.