Identification of biomarkers and exploration of mechanisms for atopic dermatitis based on transcriptome and scRNA-seq data analysis.

Atopic dermatitis (AD) is one of the most common chronic inflammatory skin diseases with complex pathogenesis and no effective treatment. This study aims to use bioinformatics methods to identify biomarkers and explore the mechanism for AD. We performed differential expression analysis based on transcriptome datasets GSE16161 and GSE32924. Next, the differentially expressed genes (DEGs) were subjected to Kyoto Encyclopedia of Genes and Genomes enrichment analysis. After integrating PPI network obtained from STRING database, we screened modular genes and identified candidate key genes by the MCC and degree algorithms. We selected genes with strong ROC performance and consistent expression levels as key genes, and constructed a nomogram to assess their potential as AD biomarkers. Finally, by analyzing the scRNA-seq dataset GSE180885, we identified the key cells associated with the key genes and conducted pseudotime analysis based on these key cells to explore the pathogenic mechanisms of AD. The results showed that 618 DEGs were identified and some important pathways, including Cytokine-Cytokine Receptor Interaction, Cell Cycle, Cell Adhesion Molecules and Calcium Signaling Pathway were screened out. Seven key genes were identified and they were CCNA2, CCNB1, KIF2C, CEP55, MELK, CDC20, and CCNB2. The nomogram analysis suggested that these key genes had the potential to serve as biomarkers for AD. Through scRNA-seq data analysis, we identified 9 cell subpopulations, with keratinocytes were identified as the key cells, and 6 out of 7 key genes showed significant expression in keratinocytes. Pseudotime analysis revealed that DEGs in keratinocytes played a vital role in the cellular differentiation process of AD. We successfully identified CCNA2, CCNB1, KIF2C, CEP55, MELK, CDC20, and CCNB2, as potential biomarkers for AD through transcriptomic and scRNA-seq data analysis.