DysPIA

DysPIA identifies dysregulated pathways by integrating Correlation by Individual Level Product and a combined gene-pair enrichment background to detect pathway-level dysregulation in biological datasets.

Key Features:

• Correlation by Individual Level Product: Incorporates Correlation by Individual Level Product to capture gene regulatory relationships within pathways rather than treating pathways solely as gene sets.
• Combined gene-pair background: Performs enrichment analysis using a constructed combined gene-pair background that is more comprehensive than backgrounds used in Edge Set Enrichment Analysis.
• High performance metrics: Achieved Area Under the Curve (AUC) values between 0.9584 and 0.9896 in simulation studies and outperformed existing methods in identifying causal pathways, including analyses involving p53 mutations.
• Computational speed: Demonstrated substantially faster execution, approximately 1,700–8,000 times quicker than other methods when performing 10,000 permutations.
• Biological validation: Identified biologically significant dysregulated pathways in breast cancer relapse and subtype datasets that could be corroborated by existing literature.

Scientific Applications:

• Cancer pathway analysis: Detection and prioritization of dysregulated pathways in cancer datasets, including analyses related to p53 mutation contexts and breast cancer relapse/subtype studies.
• Genomic and translational research: Application to genomic studies and investigations aimed at understanding pathway-level dysregulation for translational or personalized medicine research.

Methodology:

Uses Correlation by Individual Level Product for per-sample correlation assessment, constructs a combined gene-pair background for enrichment analysis, and employs permutation testing (e.g., 10,000 permutations) for significance assessment.