Ant colony optimization for the identification of dysregulated gene subnetworks from expression data

Area of Science:

• Bioinformatics
• Systems Biology
• Computational Biology

Background:

• High-throughput technologies offer deep insights into pathway perturbations for identifying molecular targets and disease treatments.
• Classical analyses of gene expression data overlook gene interconnections crucial for cellular processes.
• Incorporating gene interaction networks aids in detecting gene modules underlying complex phenotypes.

Purpose of the Study:

• To develop a novel heuristic method for gene-level scoring and module identification in gene interaction networks.
• To overcome limitations of existing methods, such as radius-based restrictions and statistical bias towards large modules.
• To identify dysregulated gene modules associated with specific diseases by analyzing network structures.

Main Methods:

• Proposed a heuristic method inspired by Ant Colony Optimization (ACO).
• Implemented a distance-based search with penalties, avoiding radius constraints.
• Applied gene-level scoring and module identification within gene interaction networks.

Main Results:

• Tested the approach on three neurodegenerative disease datasets (Alzheimer's, Parkinson's, Huntington's).
• Demonstrated superior stability and meaningful enrichment results compared to existing methods.
• Achieved consistent performance across datasets with varying case-control proportions and sample sizes.

Conclusions:

• The network-based gene expression analysis successfully identifies disease-associated dysregulated gene modules.
• The ACO-inspired heuristic provides a stable and effective method for prioritizing relevant gene modules.
• The developed tool is publicly available for broader research application.