Evaluation of Pathway Activation for a Single Sample Toward Inflammatory Bowel Disease Classification

Area of Science:

• Computational biology
• Systems biology
• Bioinformatics

Background:

• Similar complex diseases often present with overlapping clinical symptoms, leading to misdiagnosis and mistreatment.
• Existing network-based prediction methods lack robustness and reproducibility.
• Pathway-based methods offer interpretability but often fail to characterize single-sample disease states.

Purpose of the Study:

• To develop a novel framework, Pathway Activation for Single Sample (PASS), for accurate classification of similar complex diseases.
• To enable precise diagnosis and identification of pathway biomarkers for individual patients.
• To explore disease-specific pathway alterations in single samples.

Main Methods:

• PASS quantifies pathway perturbation by assessing edge alterations and gene statistics within single samples.
• A novel method, AUCpath, evaluates pathway activation based on gene interactions and statistics.
• The framework was applied to differentiate between subtypes of inflammatory bowel disease (IBD).

Main Results:

• PASS demonstrated effectiveness in recognizing differences between similar complex diseases.
• The framework successfully applied to inflammatory bowel disease subtypes.
• PASS features show potential as interpretable indicators for precise diagnosis.

Conclusions:

• PASS provides a robust and interpretable method for single-sample disease state characterization.
• The framework facilitates the discovery of pathway biomarkers for complex diseases.
• PASS aids in understanding the biological mechanisms underlying disease variations.