Construction and validation of the diagnostic model of keloid based on weighted gene co-expression network analysis (WGCNA) and differential expression analysis

Abstract

Keloid is a disease that seriously affects the aesthetic appearance of the body. In contrast to normal skin or hypertrophic scars, keloid tissue extends beyond the initial site of injury. Patients may complain of pain, itching, or burning. Although multiple treatments exist, none is uniformly successful. Genetic advances have made it possible to explore differences in gene expression between keloids and normal skin. Identifying the biomarker for keloid is beneficial to the mechanism exploration and treatment development of keloid. In this study, we identified seven genes with significant differences in keloids through weighted gene co-expression network analysis(WGCNA) and differential expression analysis. Then, by the Lasso regression, we constructed a keloid diagnostic model using five of these genes. Further studies found that keloids could be divided into high-risk and low-risk groups by this model, with differences in immunity, m6A methylation, and pyroptosis. Finally, we verified the accuracy of the diagnostic model in clinical RNA-sequencing data.