pathVar

pathVar identifies pathways and gene sets exhibiting significantly different gene expression variability relative to a reference gene set to enable functional interpretation of transcriptional regulation.

Key Features:

• Functional interpretation of variability: Provides functional interpretation of changes in gene expression variability at the pathway and gene set levels to inform mechanistic models of transcriptional regulation.
• Statistical methodology: Assesses pathway-level variability using a multinomial exact test with an option for an asymptotic Chi-squared test for computational efficiency.
• Versatility across technologies and settings: Applies to gene expression data from any technology platform and supports analyses of single phenotypic groups or two-group comparisons.
• Benchmarking and validation: Validated across multiple diseases, species, and sample types and benchmarked against average-expression analyses and Gene Set Enrichment Analysis (GSEA).
• Guidance on variability statistics: Provides recommendations for selecting variability statistics informed by simulations and analyses of real datasets.
• Clustering of genes by variability: Clusters genes within pathways based on their variability levels to identify significant categories or clusters contributing to phenotypic changes.

Scientific Applications:

• Developmental Biology: Identifies pathways with altered gene expression variability to highlight regulatory genes involved in developmental processes.
• Cancer Genomics: Distinguishes cancerous and normal tissues by pathway-level variability analysis to reveal potential therapeutic targets.
• Neurological Disease Research: Analyzes variability in gene expression pathways to investigate regulatory mechanisms underlying neurological disorders.

Methodology:

Clusters genes within pathways based on variability levels and tests pathway-level variability using a multinomial exact test or an asymptotic Chi-squared test.