Background Keloids, while considered a benign skin condition, can lead to significant discomfort. The exact mechanisms behind their development remain obscure. Methods The objective of the current investigation was to find key genes involved in keloid formation and explore the interactions between mitophagy and the immune microenvironment. Datasets GSE145725 and GSE7890 were sourced from the Gene Expression Omnibus (GEO) database. Differentially expressed genes (DEGs) together with module genes were found and subjected to functional enrichment analysis. To investigate mitophagy‐related hub genes and their role in keloid pathology, machine learning (ML) techniques, including least absolute shrinkage and selection operator (LASSO) regression and random forest, were applied. A nomogram was created using the found hub genes. The immune landscape was analyzed using single‐sample gene set enrichment analysis (ssGSEA). Hub gene protein expression and immune cell infiltration in clinical samples were evaluated using Western blot analysis and immunofluorescence labeling. Results Four hub genes related to mitochondrial autophagy—ARHGAP29, PLXDC2, YPEL3, and ENTPD7—were identified through ML. A diagnostic nomogram model constructed from these four genes demonstrated high reliability and accuracy. The immune microenvironment varied between keloid and control groups, with Spearman correlation analysis revealing both positive and negative associations between the hub genes and infiltrating immune cells. Additionally, reduced protein expression of ARHGAP29 and YPEL3, along with increased infiltration of CD8+ T cells, was observed in keloid tissues. Conclusion This work offers a thorough examination of immune infiltration and mitophagy in keloid development.
Wang et al. (Thu,) studied this question.