qRT‒PCR and WB data.

BackgroundUlcerative colitis (UC) is a chronic nonspecific inflammatory bowel disease of unknown etiology that is associated with a significant risk of progression to colorectal cancer. The aim of this study was to systematically identify hypoxia- and mitophagy-related molecular signatures associated with UC, thereby providing novel insights into disease mechanisms and therapeutic strategies.MethodsA comprehensive analytical framework integrating differential expression analysis and functional enrichment assessment was employed to systematically characterize dysregulated mitophagy-related genes (MRGs) and hypoxia-related genes (HRGs) in UC and their associated pathogenic pathways. We employed two advanced machine learning methods, support vector machine with recursive feature elimination (SVM-RFE) and least absolute shrinkage and selection operator (LASSO), to evaluate diagnostic models validated by receiver operating characteristic (ROC) curves and optimize feature selection. These results were verified by basic experiments. We subsequently analyzed immune cell infiltration to clarify the interaction between mitophagy/hypoxia and immunological disorders in UC pathogenesis. Finally, mRNA–transcription factor (TF) and mRNA–miRNA regulatory networks were constructed, revealing intricate molecular crosstalk among hub genes through systematic bioinformatic analyzes.ResultsAfter validation with two machine learning approaches, two pivotal biomarkers (CD55 and CPT1A) with diagnostic potential were rigorously selected. ROC curve analysis revealed the superior diagnostic efficacy of these key genes, confirming their clinical discriminative capacity. Experimental verification confirmed these findings. Notably, subsequent immune profiling revealed significant upregulation of multiple immune cell populations in the high-risk UC subgroup. Furthermore, the expression of diagnostic biomarkers was significantly correlated with dynamic changes in immune cell infiltration, suggesting that these biomarkers play immunomodulatory roles in UC progression. Finally, mRNA–miRNA and mRNA–TF regulatory network analyzes revealed complex interactions.ConclusionsWe elucidated the relationship between UC and hypoxia/mitophagy and identified potential diagnostic biomarkers. This study provides a reference for the future development of targeted treatment strategies to improve diagnostic and therapeutic protocols for UC.背景：溃疡性结肠炎（Ulcerative colitis, UC）是一种病因未明的慢性非特异性炎症性肠病，患者进展为结直肠癌的风险显著升高。本研究旨在系统性筛选与溃疡性结肠炎相关的缺氧及线粒体自噬相关分子特征，以期为阐明疾病发病机制、开发治疗策略提供全新视角。 方法：本研究采用整合差异表达分析与功能富集评估的综合分析框架，系统性刻画溃疡性结肠炎患者中失调的线粒体自噬相关基因（mitophagy-related genes, MRGs）与缺氧相关基因（hypoxia-related genes, HRGs）及其潜在致病通路。研究运用两种先进机器学习方法——带递归特征消除的支持向量机（support vector machine with recursive feature elimination, SVM-RFE）与最小绝对收缩和选择算子（least absolute shrinkage and selection operator, LASSO）构建诊断模型，通过受试者工作特征（receiver operating characteristic, ROC）曲线验证模型性能并优化特征筛选流程。上述研究结果均通过基础实验予以验证。随后，本研究分析了免疫细胞浸润情况，以阐明线粒体自噬/缺氧与溃疡性结肠炎发病过程中免疫紊乱的相互作用。最后，构建了mRNA-转录因子（transcription factor, TF）及mRNA-微小RNA（miRNA）调控网络，通过系统性生物信息学分析揭示核心基因间复杂的分子串扰机制。 结果：经两种机器学习方法验证，本研究严格筛选出两个具有诊断潜力的关键生物标志物：CD55与CPT1A。ROC曲线分析显示，这些关键基因具备优异的诊断效能，证实了其临床区分能力。实验验证进一步确认了上述发现。值得注意的是，后续免疫细胞谱分析显示，高风险溃疡性结肠炎亚组内多种免疫细胞群体的表达水平显著上调。此外，诊断生物标志物的表达水平与免疫细胞浸润的动态变化显著相关，提示这些标志物在溃疡性结肠炎进展中发挥免疫调节作用。最后，mRNA-miRNA及mRNA-TF调控网络分析揭示了复杂的分子相互作用。 结论：本研究阐明了溃疡性结肠炎与缺氧/线粒体自噬之间的关联，并筛选出潜在的诊断生物标志物。本研究为未来开发靶向治疗策略、优化溃疡性结肠炎的诊疗方案提供了参考依据。