Identification of potential biomarkers for ulcerative colitis based on multidimensional data analysis and investigation of the mechanisms of therapeutic effects

OBJECTIVE: Based on multidimensional data analysis, potential biomarkers for ulcerative colitis were screened, and the effects of curcumin chitosan microspheres on the expression of key targets in ulcerative colitis and their role in alleviating inflammation were observed. METHODS: Potential biomarkers related to UC progression were identified through differential expression analysis (using the limma package, with |log2FC| > 1 and corrected P < 0.05), weighted gene coexpression network analysis, and three machine learning algorithms (LASSO, random forest, SVM-RFE) on the basis of the datasets GSE107499 and GSE87473 in the GEO database. The core genes selected through this process were externally validated via the independent dataset GSE47908. A UC mouse model was constructed using 6-8-week-old C57BL/6 mice, and CCM was applied to the UC mice. Colonic tissues were collected for pathological examination and determination of inflammatory factor levels. Changes in the protein expression of the core genes were detected via immunohistochemistry. RESULTS: After rigorous screening and external validation, four core genes were finally identified: FCN3, FGR, HSD11B1 and PIM2. CCM treatment improved pathological damage and inflammatory factor levels in the colon tissue of UC mice, and the protein expression levels of FCN3, FGR, HSD11B1 and PIM2 in the colon tissue were inhibited. CONCLUSION: Four potential targets of UC, namely, FCN3, FGR, HSD11B1, and PIM2, were identified. CCM may improve the degree of the UC inflammatory response by downregulating the expression of key genes.