Identification and characterization of ion channel-related signatures in ulcerative colitis: an integrated multi-omics and machine learning analysis

Background Ulcerative colitis (UC) is a chronic, multifactorial inflammatory bowel disease characterized by persistent mucosal inflammation and immune dysregulation. Human ion channel genes (HICGs) are involved in maintaining cellular homeostasis and regulating immune function, but their expression patterns and diagnostic value in UC have not been systematically investigated. This study aimed to identify UC-associated HICGs and evaluate their potential as diagnostic biomarkers using an integrative multi-omics approach. Methods Multiple transcriptomic datasets were integrated with HICGs to identify UC-associated genes. Differential expression analysis, weighted gene co-expression network analysis (WGCNA), and machine learning were used to identify biomarkers and construct a diagnostic model validated in independent cohorts. Immune infiltration, single-cell RNA sequencing (scRNA-seq), spatial transcriptomics, cell–cell communication, virtual screening, and DSS-induced colitis validation were performed to characterize their biological and immune relevance. Results By integrating differential expression analysis, WGCNA, and machine learning, four core HICGs ( TRPM6 , FXYD3 , CNGA1 , and KCNJ15 ) were identified. The diagnostic model showed good performance in the training cohort (AUC = 0.979) and independent cohorts (AUC range: 0.864–0.981). TRPM6 , FXYD3 , and CNGA1 were downregulated in inflamed UC tissues, whereas KCNJ15 was upregulated. Their expression patterns were associated with histopathological mucosal status and disease activity. Immune infiltration analysis showed that TRPM6 , FXYD3 , and CNGA1 were negatively correlated with pro-inflammatory immune populations, whereas KCNJ15 showed the opposite pattern. Single-cell RNA sequencing demonstrated that TRPM6 , FXYD3 , and CNGA1 were mainly expressed in intestinal epithelial cells and reduced in UC, while KCNJ15 was enriched in mononuclear phagocytes and neutrophils and increased in UC. Spatial transcriptomics confirmed reduced expression of TRPM6 , FXYD3 , and CNGA1 in UC. RT-qPCR and immunohistochemistry (IHC) further confirmed the expression changes of these four HICGs in DSS-induced colitis, which were consistent with those observed in clinical UC samples. Conclusion TRPM6 , FXYD3 , CNGA1 , and KCNJ15 were identified as HICG-based biomarkers with diagnostic potential for UC. Multi-omics analyses revealed their cell-type-specific expression patterns and associations with immune microenvironment remodeling, highlighting their potential roles in UC pathogenesis.