A novel multivariate framework for functional gene networks enrichment analysis

Area of Science:

• Computational Biology and Bioinformatics
• Systems Biology
• Genomics and Network Medicine

Background:

• Gene network analysis is crucial for understanding disease mechanisms but interpreting large networks is challenging.
• Current methods often reduce complex network features to a single score, masking important multivariate characteristics.
• This limitation hinders the differentiation of individual network component contributions in biological and disease processes.

Purpose of the Study:

• To introduce and validate a novel computational strategy, Multivariate Framework for Functional Gene Network Enrichment Analysis (mFGNA).
• To incorporate diverse graph-theoretical network features for a comprehensive analysis of gene networks.
• To enable effective functional pathway discovery and identify potential translational targets in disease contexts.

Main Methods:

• Developed mFGNA to integrate node properties, edge connectivity, interaction strengths, and expression levels.
• Utilized a gene-level permutation strategy for robust statistical inference and reduced computational complexity.
• Validated mFGNA using extensive Monte Carlo simulations on undirected and directed gene networks across diverse pathway settings.

Main Results:

• mFGNA effectively preserved multidimensional network characteristics, capturing complex gene network rewiring across phenotypic states.
• Simulations demonstrated superior performance of mFGNA over existing approaches in various pathway enrichment scenarios.
• Application to cancer cell lines identified significant immune pathway dysregulation in pancreatic and non-small cell lung cancers, with specific modules dominated by HLA class II genes.

Conclusions:

• mFGNA provides a powerful tool for functional pathway discovery in complex gene networks.
• The framework offers mechanistic insights into immune remodeling in tumors, highlighting potential diagnostic and therapeutic targets.
• mFGNA enables effective differentiation of network components, advancing disease research and translational applications.