Mining pathway associations for disease-related pathway activity analysis based on gene expression and methylation...

Area of Science:

• Genomics and Bioinformatics
• Systems Biology
• Cancer Research

Background:

• Identifying genetic markers is crucial for diagnosing and treating complex diseases.
• Previous studies focused on gene or gene-set markers, which may not capture complex genetic interactions.
• Distinctive associations among active pathways (pathway-sets) are investigated using gene expression and/or methylation data.

Purpose of the Study:

• To investigate distinctive associations among active pathway-sets in both case and control samples.
• To identify novel genetic markers for disease signatures beyond traditional gene or gene-set markers.
• To develop a method for visualizing differential pathway associations between disease states.

Main Methods:

• Pathway-sets were identified by finding sets of pathways frequently active together in samples.
• Gene-set enrichment analysis was used to identify significant (active) pathways from gene expression or methylation data.
• Association rule mining was applied to active pathways to find distinctive pathway-sets for case and control groups, forming a pathway activity network (PAN).

Main Results:

• Distinctive pathway-set signatures were identified for breast cancer and uterine leiomyoma cancer using two public datasets.
• The pathway activity network (PAN) effectively visualized differential pathway associations between case and control samples.
• Pathway-set markers demonstrated superior or comparable performance to gene or gene-set markers in disease classification.

Conclusions:

• Pathway-set markers offer a powerful new approach for disease classification, providing deeper biological insights.
• The pathway activity network (PAN) aids in understanding complex pathway interactions relevant to disease.
• This methodology advances the discovery of genetic markers for improved disease diagnosis, treatment, and prognosis.