Development and Experimental Validation of Machine Learning-Based Disulfidptosis-Related Ferroptosis Biomarkers in Inflammatory Bowel Disease
Abstract
Background: Inflammatory bowel disease (IBD) is a chronic inflammatory condition of the gastrointestinal tract, defined by intestinal epithelial cell death. While ferroptosis and disulfidptosis have been linked to IBD pathogenesis, the functional significance of disulfidptosis-related ferroptosis genes (DRFGs) in this disease remains poorly characterized. This investigation sought to pinpoint DRFGs as diagnostic indicators and clarify their mechanistic contributions to IBD progression. Methods: Four IBD datasets (GSE65114, GSE87473, GSE102133, and GSE186582) from the GEO database were integrated to identify differentially expressed genes (DEGs) (|log2FC| > 0.585, adj. p < 0.05). A Pearson correlation analysis was used to link disulfidptosis and ferroptosis genes, followed by machine learning (LASSO and RF) to screen core DRFGs. The immune subtypes and single-cell sequencing (GSE217695) results were analyzed. A DSS-induced colitis Mus musculus (C57BL/6) model was used for validation. Results: Transcriptomic profiling identified 521 DEGs, with 16 defined as DRFGs. Nine hub genes showed diagnostic potential (AUC: 0.71-0.91). Functional annotation demonstrated that IBD-associated genes regulate diverse pathways, with a network analysis revealing their functional synergy. The PPI networks prioritized DUOX2, NCF2, ACSL4, GPX2, CBS, and LPCAT3 as central hubs. Two immune subtypes exhibited divergent DRFG expression. Single-cell mapping revealed epithelial/immune compartment specificity. The DSS-induced murine colitis model confirmed differential expression patterns of DRFGs, with concordant results between qRT-PCR and RNA-seq, emphasizing their pivotal regulatory roles in disease progression and potential for translational application. Conclusions: DRFGs mediate IBD progression via multi-signal pathway regulation across intestinal cell types, demonstrating diagnostic and prognostic potential.