Identification of cuproptosis-associated subtypes and signature genes for diagnosis and risk prediction of Ulcerative colitis based on machine learning

BACKGROUND: Ulcerative colitis (UC) is a chronic and debilitating inflammatory bowel disease that impairs quality of life. Cuproptosis, a recently discovered form of cell death, has been linked to many inflammatory diseases, including UC. This study aimed to examine the biological and clinical significance of cuproptosis-related genes in UC. METHODS: Three gene expression profiles of UC were obtained from the Gene Expression Omnibus (GEO) database to form the combined dataset. Differential analysis was performed based on the combined dataset to identify differentially expressed genes, which were intersected with cuproptosis-related genes to obtain differentially expressed cuproptosis-related genes (DECRGs). Machine learning was conducted based on DECRGs to identify signature genes. The prediction model of UC was established using signature genes, and the molecular subtypes related to cuproptosis of UC were identified. Functional enrichment analysis and immune infiltration analysis were used to evaluate the biological characteristics and immune infiltration landscape of signature genes and molecular subtypes. RESULTS: Seven signature genes (ABCB1, AQP1, BACE1, CA3, COX5A, DAPK2, and LDHD) were identified through the machine learning algorithms, and the nomogram built from these genes had excellent predictive performance. The 298 UC samples were divided into two subtypes through consensus cluster analysis. The results of the functional enrichment analysis and immune infiltration analysis revealed significant differences in gene expression patterns, biological functions, and enrichment pathways between the cuproptosis-related molecular subtypes of UC. The immune infiltration analysis also showed that the immune cell infiltration in cluster A was significantly higher than that of cluster B, and six of the characteristic genes (excluding BACE1) had higher expression levels in subtype B than in subtype A. CONCLUSIONS: This study identified several promising signature genes and developed a nomogram with strong predictive capabilities. The identification of distinct subtypes of UC enhances our current understanding of UC’s underlying pathogenesis and provides a foundation for personalized diagnosis and treatment in the future.