Screening and experimental study of potential biomarkers for ulcerative colitis based on weighted gene co-expression network analysis and machine learning

Zepeng Chen ^1,2, Xingchen Wang ^1,2, Changfang Xiao ¹

⁰Department of Anorectal Surgery, LongHua Hospital, Shanghai University of Traditional Chinese Medicine, Shanghai, 200032, China.

Journal of Translational Medicine +

|

October 1, 2025

View abstract on PubMed

Summary

Poly(ADP-ribose) polymerase family member 8 (PARP8) is identified as a potential biomarker for ulcerative colitis (UC). This finding, derived from integrated bioinformatics and machine learning, suggests PARP8 plays a role in UC pathogenesis through immune pathways.

Area Of Science

Immunology
Genomics
Bioinformatics

Background

Ulcerative colitis (UC) is a chronic inflammatory bowel disease with unclear pathogenesis.
It affects the intestinal mucosa and submucosa, causing continuous inflammation.

Purpose Of The Study

To identify novel ulcerative colitis (UC) biomarkers.
Integrate weighted gene co-expression network analysis (WGCNA) with machine learning for biomarker discovery.
Validate potential biomarkers in an experimental UC mouse model.

Main Methods

Systematic querying of the Gene Expression Omnibus database (GSE87466 dataset).
Identification of differentially expressed genes (DEGs) and enrichment analyses.
Application of WGCNA, LASSO, and SVM-RFE for biomarker screening.
Validation through in vivo animal experiments.

Main Results

1,097 DEGs were identified; WGCNA highlighted a turquoise module strongly associated with UC.
Poly(ADP-ribose) polymerase family member 8 (PARP8) was identified as a key biomarker by LASSO and SVM-RFE.
PARP8 expression correlated with increased neutrophils, M1 macrophages, and activated CD4+ T cells in UC patients.
Elevated PARP8 expression was confirmed in a mouse model of UC.

Conclusions

PARP8 may contribute to ulcerative colitis (UC) pathogenesis via immune-related pathways.
PARP8 shows promise as a diagnostic and predictive biomarker for UC.
Further research into PARP8's role in immune dysregulation in UC is warranted.

Keywords:

Differentially expressed genes Machine learning Ulcerative colitis Weighted gene co-expression network analysis