PathExt: a general framework for path-based mining of omics-integrated biological networks

Area of Science:

• Genomics and Bioinformatics
• Systems Biology
• Computational Biology

Background:

• Transcriptome analysis commonly focuses on differentially expressed genes (DEGs) to identify phenotype-related genes.
• This approach overlooks the network interactions of genes and proteins crucial for phenotype emergence.
• Practical limitations like insufficient samples or lack of controls hinder robust DEG identification.

Purpose of the Study:

• To introduce PathExt, a computational tool for identifying differentially active paths in omics-integrated biological networks.
• To enable the extraction of relevant biological context, genes, and processes through a TopNet sub-network.
• To provide a method for analyzing biological systems with limited samples or absent controls.

Main Methods:

• PathExt identifies differentially active paths in omics-integrated networks, with or without a control group.
• It extracts a TopNet sub-network representing key genes and processes.
• The tool is validated using transcriptomic data for Mycobacterium tuberculosis drug response and human tissue-specific gene analysis.

Main Results:

• PathExt successfully identifies characteristic genes and pathways, even with single-sample data.
• The tool can analyze biological systems lacking appropriate controls.
• Case studies demonstrated PathExt's utility in characterizing drug responses and tissue-specific biological processes.

Conclusions:

• PathExt is a versatile tool for prioritizing context-relevant genes and pathways in any omics-integrated biological network.
• It overcomes limitations of DEG analysis, particularly in scenarios with scarce data or no controls.
• The tool facilitates a deeper understanding of biological systems and phenotype underpinnings.