Identifying colon cancer risk modules with better classification performance based on human signaling network.

Area of Science:

• Oncology
• Bioinformatics
• Systems Biology

Background:

• Signaling transduction and biological pathways are often disrupted in cancer.
• Gene expression profiles can reveal disease-specific variations.
• Understanding molecular differences between normal and tumor tissues is crucial for colon cancer research.

Purpose of the Study:

• To identify molecular mechanisms underlying colon cancer by analyzing differences between normal and tumor samples from a modular perspective.
• To develop a method for selecting risk modules associated with tumor conditions using integrated biological networks and gene expression data.

Main Methods:

• Integrated a weighted human signaling network with gene expression profiles.
• Utilized selected risk modules as classification features.
• Evaluated classification performance for distinguishing normal/tumor samples and tumor stages.

Main Results:

• The proposed method, using risk modules, demonstrated superior classification performance compared to other approaches.
• A specific risk module for colon cancer effectively differentiated between normal and tumor samples and across different tumor stages.
• Genes within this key module were primarily associated with the positive regulation of cell proliferation and showed a strong link to colon cancer.

Conclusions:

• The identified risk modules, particularly the one linked to cell proliferation, show potential as biomarkers for colon cancer detection and staging.
• These findings suggest that the genes within the identified module may represent crucial risk factors for colon cancer development.
• A modular approach integrating signaling networks and gene expression is effective for uncovering disease mechanisms and identifying potential therapeutic targets.