Advances in Inflammatory Bowel Disease Diagnostics: Machine Learning and Genomic Profiling Reveal Key Biomarkers for...

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Inflammatory Bowel Disease (IBD) presents complex diagnostic challenges.
• Identifying reliable biomarkers is crucial for early detection and personalized management of IBD.
• Current diagnostic methods may lack the precision needed for effective patient stratification.

Purpose of the Study:

• To identify novel gene biomarkers and molecular signatures for Inflammatory Bowel Disease (IBD) using high-throughput data analysis.
• To develop a machine learning model for accurate IBD diagnosis based on gene expression profiles.
• To validate the diagnostic potential of identified biomarkers in independent patient cohorts.

Main Methods:

• Utilized microarray gene expression data from 172 IBD patients and 22 healthy controls (GSE75214).
• Applied machine learning techniques and feature selection to identify Differentially Expressed Gene (DEG) biomarkers.
• Validated findings using independent datasets (GSE36807 and GSE10616) and performed functional/pathway enrichment analysis.

Main Results:

• Identified six significant DEG biomarkers (VWF, IL1RL1, DENND2B, MMP14, NAAA, PANK1) with high diagnostic potential for IBD.
• The Random Forest model achieved excellent performance (accuracy, F1-score, AUC > 0.98) in the primary dataset.
• Validated biomarkers demonstrated favorable performance in independent datasets (accuracy: 0.841, F1-score: 0.734, AUC: 0.887).
• DENND2B and PANK1 were identified as novel IBD biomarkers.

Conclusions:

• The identified gene biomarkers, particularly DENND2B and PANK1, offer significant potential for the early detection and personalized treatment of IBD.
• Machine learning models based on these biomarkers show high accuracy and reliability, validated across multiple datasets.
• These findings contribute valuable insights into IBD pathogenesis and pave the way for improved diagnostic tools and therapeutic strategies.