Identification of circadian rhythm-related biomarkers and development of diagnostic models for Crohn's disease using machine learning algorithms.

Abstract

The global rise in Crohn's Disease (CD) incidence has intensified diagnostic challenges. This study identified circadian rhythm-related biomarkers for CD using datasets from the GEO database. Differentially expressed genes underwent Weighted Gene Co-Expression Network Analysis, with 49 hub genes intersected from GeneCards data. Diagnostic models were constructed using machine learning algorithms, and biologic therapy efficacy was predicted with advanced regression techniques. Single-cell sequencing showed high gene expression in stem cells, immune, and endothelial cells, with validation confirming significant differences between CD patients and controls. These findings suggest circadian rhythm-related genes as promising diagnostic biomarkers for CD.