Identification of DYRK2 and TRIM32 as keloids programmed cell death-related biomarkers: insights from bioinformatics and machine learning in multiple cohorts

This study aims to explore the expression patterns and mechanisms of programmed cell death-related genes in keloids and identify molecular targets for early diagnosis and treatment. We first explored the expression, immune, and biological function profiles of keloids. Using various machine learning methods, two key genes, DYRK2 and TRIM32, were identified, with ROC curves demonstrating their diagnostic potential. Further analyses, including GSEA, immune cell profiling, competing endogenous RNA network, and single-cell analysis, revealed their mechanism of action and regulatory network. Finally, SB-431542 was identified as a potential therapeutic agent for keloids through CMap and molecular docking.