Disease-related gene module detection based on a multi-label propagation clustering algorithm.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Analyzing gene expression data is crucial for understanding complex disease mechanisms.
• Multi-label propagation algorithm (MLPA) is a common method for gene module detection but struggles with loosely connected genes and lacks demonstrated biological significance.
• Existing methods often overlook the importance of pathogenic information and hierarchical gene structures.

Purpose of the Study:

• To develop an improved gene module detection algorithm for analyzing gene expression data.
• To address the limitations of MLPA, particularly its performance with loosely connected disease-related genes.
• To enhance the biological relevance and accuracy of detected gene modules, using Huntington's disease as a case study.

Main Methods:

• Designed a Double Label Propagation Clustering Algorithm (DLPCA) building upon the MLPA framework.
• Integrated both category labels and novel pathogenic labels to guide the multi-label propagation process.
• Utilized pathogenic labels containing disease gene information and hierarchical gene expression structures.

Main Results:

• DLPCA demonstrated superior performance compared to conventional gene-clustering algorithms.
• The algorithm effectively identified disease-related gene modules with improved accuracy and biological significance.
• Experimental results validated the enhanced capabilities of DLPCA in analyzing complex disease phenotypes.

Conclusions:

• DLPCA offers a significant advancement in detecting disease-related gene modules from gene expression data.
• The incorporation of pathogenic labels enhances the robustness and biological interpretability of clustering results.
• This approach shows promise for future studies on complex diseases and gene interaction mechanisms.