Background: Hidradenitis suppurativa (HS) is a chronic auto-inflammatory skin condition characterized by nodules, abscesses, and fistulae in skin folds. The underlying pathogenesis of HS remains unclear, and effective therapeutic drugs are limited.
Methods: We acquired mRNA expression profiles from the Gene Expression Omnibus (GEO) database and conducted differential expression analysis between control and HS samples using R software. Four machine learning algorithms (SVM, RF, ANN, and lasso) and WCGNA were utilized to identify feature genes. GO, KEGG, Metascape, and GSVA were utilized for the enrichment analysis. CIBERSORT and ssGSEA were employed to analyze immune infiltration.
Results: A total of 29 DEGs were identified, with the majority showing up-regulation in HS. Enrichment analysis revealed their involvement in immune responses and cytokine activities. KEGG analysis highlighted pathways such as IL-17 signaling, rheumatoid arthritis, and TNF signaling in HS. Immune infiltration analysis revealed the predominant presence of neutrophils, monocytes, and CD8 T cells. Machine learning algorithms and WCGNA identified KYNU as a feature gene associated with HS. We have also identified 59 potential drugs for HS based on the DEGs. Additionally, ceRNA network analysis identified the MUC19_hsa-miR-382-5p_KYNU pathway as a potential regulatory pathway.
Conclusions: KYNU emerged as a feature gene associated with HS, and the ceRNA network analysis identified the MUC19_hsa-miR-382-5p_KYNU pathway as a potential regulator.
Keywords: Hidradenitis suppurativa; WCGNA; bioinformatics; biomarker; machine learning.