Integrative biomarker discovery and immune profiling for ulcerative colitis: a multi-methodological approach

Area of Science:

• Gastroenterology
• Immunology
• Computational Biology

Background:

• Ulcerative colitis (UC) is an autoimmune condition requiring better diagnostic tools and understanding of its immune basis.
• Current diagnostic and therapeutic strategies for UC can be improved through biomarker discovery and immune profiling.

Purpose of the Study:

• To identify novel biomarkers for ulcerative colitis (UC).
• To develop a diagnostic model for UC using machine learning.
• To investigate the immune cell landscape and patterns in UC.

Main Methods:

• Utilized bulk and single-cell sequencing data from the Gene Expression Omnibus (GEO).
• Applied differential analysis, Weighted Gene Co-expression Network Analysis (WGCNA), Protein-Protein Interaction (PPI), LASSO, Random Forest (RF), and Support Vector Machine-Recursive Feature Elimination (SVM-RFE) for biomarker selection.
• Developed a diagnostic model using a neural network and validated it with ROC analysis. Performed immune cell profiling with CIBERSORT.

Main Results:

• Identified eight key biomarkers: B4GALNT2, PDZK1IP1, FAM195A, REG4, MTMR11, FLJ35024, CD55, and CD44.
• The diagnostic model achieved high accuracy with AUCs of 0.984 (training) and 0.957 (validation).
• Observed increased plasma cells and CD8 T cells in UC tissues, revealing two distinct immune patterns involving T and NK cells.

Conclusions:

• Successfully identified eight novel biomarkers for ulcerative colitis (UC).
• Constructed a robust neural network-based diagnostic model for UC.
• Provided insights into the immune complexity of UC, aiding future diagnosis and treatment strategies.