Identification and Subtype Analysis of Interleukin-17-related Diagnostic Biomarkers in Atopic Dermatitis Based on WGCNA and Machine Learning
TL;DR
This integrated analysis identified IL4R and PRSS22 as key IL-17-related diagnostic biomarkers for AD, demonstrating excellent diagnostic accuracy across both training and validation cohorts, and highlighted potential regulatory mechanisms involving these biomarkers.
Abstract
Atopic dermatitis (AD) is a chronic inflammatory skin disease with heterogeneous immune dysregulation. Although interleukin-17 (IL-17) signaling is implicated in AD, IL-17-related diagnostic biomarkers and molecular subtypes remain unclear. This study aimed to identify IL-17-associated biomarkers, characterize immune infiltration and subtypes, and explore upstream regulatory mechanisms. Gene expression datasets (GSE121212, GSE6012) were acquired from the Gene Expression Omnibus database, and an IL-17-related gene set was collected from the Gene Set Enrichment Analysis website (GSEA)(http://www.gsea-msigdb.org/gsea/msigdb/search.jsp). Differential expression (limma) and Weighted Gene Co-expression Network Analysis were integrated to identify candidate genes, followed by feature selection using Least Absolute Shrinkage and Selection Operator and random forest. We evaluated immune cell infiltration by applying the CIBERSORT algorithm alongside single-sample Gene Set Enrichment Analysis. GSEA was applied to investigate underlying biological processes. AD patients were clustered into subtypes relying on IL-17 scores using ssGSEA. Our integrated analysis identified IL4R and PRSS22 as key IL-17-related diagnostic biomarkers for AD, demonstrating excellent diagnostic accuracy across both training and validation cohorts. Immune-infiltration analyses revealed altered immune-cell composition and correlations observed between the identified biomarkers and specific immune cells. Two distinct AD subtypes were identified based on IL-17 scores, exhibiting immune infiltration patterns and enriched biological pathways. A ceRNA network highlighted potential regulatory mechanisms involving these biomarkers. IL4R and PRSS22 are robust IL-17-related diagnostic biomarkers for AD, with high predictive power across cohorts. Immune infiltration profiling and IL-17 score-based subtyping reveal AD heterogeneity. These findings provide a foundation for improved diagnosis, molecular stratification, and potential therapeutic targeting in AD.