rPAC: Route based pathway analysis for cohorts of gene expression data sets

Area of Science:

• Bioinformatics
• Systems Biology
• Genomics

Background:

• High-throughput gene expression studies require robust pathway analysis for biological interpretation.
• Existing methods primarily identify perturbed pathways, lacking granular route analysis.

Purpose of the Study:

• Introduce rPAC, a novel framework for pathway route perturbation analysis.
• Enhance the deciphering of disease etiology by isolating specific pathway sections.

Main Methods:

• Decompose signaling pathways into upstream and downstream portions relative to transcription factor blocks.
• Generate disturbance scores for pathway route segments.
• Utilize summary metrics: Proportion of Significance (PS) and Average Route Score (ARS).

Main Results:

• rPAC demonstrated superior performance compared to conventional methods in simulated data.
• Identified specific pathway routes as potential cancer type signatures in TCGA epithelial cancer datasets.
• Revealed known and novel pathway routes associated with patient subgroups (e.g., age, cancer subtypes).

Conclusions:

• rPAC offers a more granular approach to pathway analysis, improving the identification of perturbed routes.
• The framework enhances the discovery of disease-specific signatures and potential etiological insights.
• rPAC shows promise in deciphering complex diseases by isolating pathway perturbations with finer granularity.