Abstract
BACKGROUND: Numerous studies have reported that dysregulation of fatty acid metabolic pathways is associated with the pathogenesis of vitiligo, in which arachidonic acid metabolism (AAM) plays an important role. However, the molecular mechanisms of AAM in the pathogenesis of vitiligo have not been clarified. Therefore, we aimed to identify the biomarkers and molecular mechanisms associated with AAM in vitiligo using bioinformatics methods. METHODS: The GSE75819 and GSE65127 datasets were used in this study as the training and validation sets, respectively, along with 58 AAM-related genes (AAM-RGs). The differentially expressed genes (DEGs) between the lesional and control groups in the training set were identified through differential expression analysis. A biomarker-based nomogram was constructed to predict the risk of vitiligo. RESULTS: 15 overlapping candidate genes were obtained between the DEGs and AAM-RGs. Machine-learning algorithms were used to identify six key genes as PTGDS, PNPLA8, FAAH, ABHD12, PTGS1, and MGLL. In both the training and validation sets, PTGDS, PNPLA8, and MGLL. In both the training and validation sets, PTGDS, PNPLA8, and MGLL were regarded as biomarkers. A nomogram based on these biomarkers showed potential for predicting the risk of vitiligo. Functional enrichment, immune cell infiltration, and regulatory network analyses were used to elucidate the molecular mechanisms. CONCLUSION: In conclusion, PTGDS, PNPLA8, and MGLL were implicated in AAM to influence the pathogenesis of vitiligo. These findings offer insights into vitiligo treatment, although further research is needed for a comprehensive understanding.