Although post-translational modifications (PTMs) significantly influence the pathogenesis of psoriasis, their specific mechanisms remain poorly understood. This study aimed to identify novel PTM-targeted therapeutic avenues for the precision treatment of psoriasis. Nine psoriasis-related datasets GSE13355, GSE14905, GSE30999, GSE41664, GSE51440, GSE53552, GSE54456, GSE117239, and GSE226244 were obtained. Consensus clustering was performed to identify psoriasis subtypes. Then, the study proceeded to pinpoint module genes by leveraging WGCNA. Additionally, the investigation also focused on uncovering hub genes by machine learning. Finally, we investigated the expression and localization of DPP3. A total of 2,141 differentially expressed genes (DEGs) were identified between the control and psoriasis samples. Enrichment analyses revealed that these DEGs are associated with inflammatory pathway involved in the progression of psoriasis. Additionally, two distinct psoriasis subclasses were identified, namely cluster.A and cluster.B. Moreover, 279 module genes related to these subtypes were identified using WGCNA. Then, utilizing the shared DEGs, cluster DEGs, key module genes, and PTM genes, 7 hub genes were obtained. Multiple machine learning models screened 7 key genes, demonstrated excellent diagnostic ability in psoriasis across multiple datasets (Average AUC = 0.901), and outperformed currently utilized diagnostic indicators specific to psoriasis. From a biological standpoint, the analysis revealed significant correlations between PTM.score and various inflammatory parameters. Notably, DPP3 emerged as an essential molecular element, exhibiting higher expression levels in psoriasis. Mechanistically, the overexpression of DPP3 enhances inflammatory activation, which in turn promotes the progression of psoriasis. This research provides genetic evidence supporting the therapeutic potential of targeting DPP3 in the treatment of psoriasis, while also elucidating its underlying pathophysiological mechanisms.
