pathwayPCA

pathwayPCA implements principal component analysis-based pathway analysis approaches for testing pathway association with binary, continuous, or survival phenotypes and extracting relevant genes to estimate pathway activities for integrative analyses such as multi-omics analysis.

Key Features:

• Pathway Association Testing: Tests pathway association with binary, continuous, or survival phenotypes.
• Gene Extraction: Extracts relevant genes in pathways using the SuperPCA and AES-PCA approaches.
• Principal Component Computation: Computes principal components based on selected genes to estimate latent variables representing pathway activities for individual subjects.
• Driver Gene Identification: Extracts relevant genes that drive pathway significance along with their corresponding data for additional in-depth analysis.
• Complex Experimental Design Support: Analyzes studies with multiple covariates and interaction effects, such as testing whether pathway association with clinical phenotype differs between male and female subjects.
• Computational Efficiency: Performs analyses with parallel computing and S4-class data objects.

Scientific Applications:

• Integrative Pathway Analysis: Uses estimated latent pathway activity variables to perform multi-omics analysis.
• Clinical Phenotype Association: Tests pathway associations with clinical phenotypes while accounting for covariates and interaction effects.
• Gene-Level Investigation: Identifies and extracts driver genes for follow-up in-depth analysis of pathway significance.

Methodology:

pathwayPCA implements the principal component analysis-based pathway analysis approaches described in Chen et al. (2008), Chen et al. (2010), and Chen (2011), including the SuperPCA and AES-PCA gene selection methods.