Abstract
INTRODUCTION: Palmitoylation influences patients with obstructive sleep apnea (OSA) by modulating amyloid-β production. However, the involvement of palmitoylation-related genes (PRGs) in OSA remains unclear. This study aims to investigate this mechanism using bioinformatics approaches. METHODS: Datasets GSE38792 and GSE135917 were retrieved from the Gene Expression Omnibus (GEO) database. Differentially expressed PRGs (DE-PRGs) were identified through differential expression analysis and weighted gene co-expression network analysis (WGCNA). Candidate genes were pinpointed using the max cluster centrality method in cytoHubba. Biomarkers were selected through machine learning algorithms, expression profiling, and ROC analysis, with diagnostic potential evaluated using a nomogram. Further insights into the role of biomarkers in OSA were provided through enrichment analysis, molecular regulatory network construction, and drug prediction. RESULTS: HIF1A and PDIA3 emerged as potential biomarkers, with the nomogram showing high predictive accuracy for OSA. Enrichment analysis revealed that HIF1A and PDIA3 were co-enriched in pathways such as "focal adhesion," "olfactory transduction," "RNA degradation," "spliceosome," and "ubiquitin-mediated proteolysis." A lncRNA-miRNA-mRNA regulatory network was constructed, featuring multiple regulatory pairs, including CYTOR-hsa-miR-1-3p-HIF1A and CYTOR-hsa-miR-1-3p-PDIA3. Drug prediction analysis identified potential compounds targeting HIF1A, such as klugine, puupehenone, and isocephaeline. CONCLUSION: HIF1A and PDIA3 were highlighted as significant potential biomarkers, providing valuable insights into the molecular mechanisms of palmitoylation in OSA and potential therapeutic targets.