Bioinformatics analysis-based mining of potential markers for inflammatory bowel disease and their immune relevance

The incidence of inflammatory bowel disease (IBD) is increasing every year and is characterized by a prolonged course, frequent relapses, difficulty in curing, and a lack of more efficacious therapeutic biomarkers. The aim of this study was to find key core genes as therapeutic biomarkers for IBD.

GSE75214 in Gene Expression Omnibus (GEO) was used as the experimental set. The genes in the top 25% of standard deviation of all samples in the experimental set were subjected to systematic weighted gene co-expression network analysis (WGCNA) to find candidate genes. Then, least absolute shrinkage and selection operator (LASSO) logistic regression was used to further screen the central genes. Finally, the validity of hub genes was verified on GEO dataset GSE179285 using "BiocManager" R package.

Twelve well-preserved modules were identified in the experimental set using the WGCNA method. Among them, five modules significantly associated with IBD were screened as clinically significant modules, and four candidate genes were screened from these five modules. Then TIMP1, GUCA2B, and HIF1A were screened as hub genes. These hub genes successfully distinguished tumor samples from healthy tissues by artificial neural network algorithm in an independent test set with an area under the working characteristic curve of 0.946 for the subjects.

IBD differentially expressed gene (DEGs) are involved in immunoregulatory processes. TIMP1, GUCA2B, and HIF1A, as core genes of IBD, have the potential to be therapeutic targets for patients with IBD, and our findings may provide a new outlook on the future treatment of IBD.