PathwayPCA: an R/Bioconductor Package for Pathway Based Integrative Analysis of Multi-Omics Data

Area of Science:

• Bioinformatics
• Statistical Genetics
• Computational Biology

Background:

• Integrative pathway analysis is crucial for understanding complex diseases by connecting molecular data to biological pathways.
• Existing methods may not fully leverage modern statistical techniques for comprehensive pathway analysis.

Purpose of the Study:

• To introduce pathwayPCA, an R/Bioconductor package for advanced integrative pathway analysis.
• To provide a flexible tool for analyzing various data types and outcomes in biological studies.

Main Methods:

• Utilizes supervised and adaptive elastic-net sparse principal component analysis.
• Applicable to continuous, binary, and survival outcomes with multiple covariates and interactions.
• Integrates gene selection, sample-specific pathway activity estimation, and visualization.

Main Results:

• pathwayPCA demonstrates superior performance over alternative methods in identifying disease-associated pathways.
• Successfully applied in case studies for gene selection, prognosis prediction, and identifying sex-specific effects in kidney cancer.
• Effectively analyzes diverse molecular data, including proteomics from large consortiums.

Conclusions:

• pathwayPCA is a powerful and versatile tool for integrative pathway analysis.
• Facilitates the interpretation of complex molecular data for a wide range of biological research.
• Empowers researchers to analyze proteomics and other omics data for enhanced biological discovery.