Machine Learning-Based Gene Prioritization Identifies Novel Candidate Risk Genes for Inflammatory Bowel Disease

Area of Science:

• Genetics and Genomics
• Immunology
• Computational Biology

Background:

• Inflammatory bowel diseases (IBD) are chronic conditions influenced by genetic, immune, and environmental factors.
• Hundreds of genes are implicated in IBD, but many more likely contribute to disease development.
• Identifying novel IBD-risk genes is crucial for understanding disease mechanisms.

Purpose of the Study:

• To develop and apply a machine learning-based gene prioritization method.
• To identify novel candidate genes associated with IBD risk.
• To enhance the understanding of IBD pathogenesis.

Main Methods:

• Collected known IBD genes from genome-wide association studies.
• Annotated genes with expression and pathway information.
• Trained a machine learning model to score and classify 16,390 genes for IBD risk.

Main Results:

• Identified immune/inflammatory responses, cell adhesion, cytokine-cytokine interaction, and sulfur metabolism pathways as relevant to IBD.
• IBD genes received significantly higher prediction scores than non-IBD genes (P < 10).
• Identified 67 novel candidate IBD-risk genes lacking prior IBD publications.

Conclusions:

• The machine learning method effectively distinguished IBD-risk genes using expression and annotation data.
• The study successfully identified novel candidate IBD-risk genes.
• These findings offer new targets for IBD research and may improve understanding of the disease.