PriPath

PriPath identifies dysregulated KEGG pathways from differential RNA-seq gene expression data using grouping, scoring, and machine learning modeling to characterize disease-associated molecular mechanisms.

Key Features:

• Integration with KEGG Pathways: Maps RNA-seq differential gene expression data onto Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways to identify affected modules or pathways.
• Machine Learning Approach: Employs a machine learning model that uses KEGG pathways as grouping functions to select the most significant pathways associated with differential gene expression.
• Disease Differentiation: Trains models to distinguish disease states from controls and has been tested on 13 diverse gene expression datasets, including datasets from different cancers and other diseases.
• Biological and Clinical Relevance: Assigns biologically and clinically relevant KEGG terms to samples based on their differential gene expression profiles.
• Comparative Performance Evaluation: Evaluates performance against other similar tools and reports top results that were manually confirmed through existing literature.

Scientific Applications:

• Medical Diagnostics: Identifies pathways dysregulated in disease states to support investigation of molecular mechanisms relevant to diagnosis.
• Precision Medicine: Supports pathway-level analyses for patient stratification and the identification of potential druggable targets based on gene expression signatures.

Methodology:

Maps RNA-seq differential gene expression onto KEGG pathways, applies grouping, scoring, and modeling with KEGG pathways as grouping functions, and uses a machine learning model trained to distinguish disease versus control samples; results were compared with other tools and top findings were manually confirmed in the literature.