Related Experiment Video
Updated: May 8, 2026
Selecting Multiple Biomarker Subsets with Similarly Effective Binary Classification Performances
Published on: October 11, 2018
Shared Biomarkers LCN2 and CXCL11 for Ulcerative Colitis and Colon Cancer: Bioinformatics Analysis and Diagnostic
Zhihao Wu1, Xiaofen Sun1, Binglong Li1
1Department of Radiology, The First Clinical College, The First Affiliated Hospital, Hainan Medical University, Haikou, 570100, People's Republic of China.
Background:
Long-term chronic inflammation is an important risk factor for colon cancer. Ulcerative colitis is a complex chronic inflammatory disease. Related studies have shown that the risk of colon cancer in patients with ulcerative colitis is 2-3 times higher than that in the general population.
Method:
Transcriptome data from GEO and TCGA databases were analyzed using RStudio. Differential expression was analyzed with "limma" and disease-related genes were identified via WGCNA. Core genes were screened by GO, KEGG, and PPI analyses, and further refined using MCC, LASSO, and RF algorithms. Expression levels and diagnostic value were evaluated via ROC curves; a disease diagnosis model was constructed. Immune cell infiltration was assessed with CIBERSORT, and GSEA analysis was performed based on gene expression.
Result:
Eighty-seven common genes were identified through differential analysis and WGCNA. Using these genes, a PPI network was built, and the top 15 genes were selected by the MCC algorithm. LASSO and RF algorithms identified LCN2 and CXCL11 as characteristic genes, highly expressed in the disease group with AUC > 0.7. The diagnosis model performed well. GO, KEGG, and GSEA analyses showed immune and inflammatory responses were important in the disease, with characteristic genes enriched in immune response and cell proliferation pathways.
Conclusion:
It was found that ulcerative colitis and colon cancer have common diagnostic markers and similar pathogenic pathways and also show similarities in the immune cell infiltration microenvironment. The disease diagnosis model constructed by combining genes is superior to the diagnostic effect of a single gene on disease.
More Related Videos
06:24Author Spotlight: Liujunzi Decoction as a Traditional Chinese Treatment for Coloproctitis Cancer
Published on: October 13, 2023
07:13Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model
Published on: April 18, 2025
