Modeling the altered expression levels of genes on signaling pathways in tumors as causal bayesian networks

Area of Science:

• Computational Biology
• Genomics
• Cancer Research

Background:

• Gene expression levels in signaling pathways are crucial for cellular function.
• Signal transduction pathways (STPs) are frequently implicated in cancer development.
• Understanding causal relationships between gene expression is key to deciphering complex diseases like cancer.

Purpose of the Study:

• To investigate if causal Bayesian networks (BNs) can model gene expression levels in STPs.
• To test the hypothesis that the causal structure of STPs is altered in cancerous tissue.
• To identify potential driver genes and therapeutic targets by analyzing these alterations.

Main Methods:

• Analyzed 22 pathways (5 breast cancer-associated, 7 other cancer-associated, 10 random) using a breast cancer gene expression dataset (529 cases, 61 controls).
• Developed separate gene expression datasets for each pathway.
• Modeled gene expression levels using causal Bayesian networks (BNs).

Main Results:

• Identified significant alterations in the causal structure of gene expression within STPs in cancer cases compared to controls.
• Demonstrated that cancer-associated STPs exhibit more altered causal structures than randomly selected pathways.
• The causal BN model effectively captured these pathway alterations.

Conclusions:

• Causal Bayesian networks are effective tools for modeling gene expression in signaling pathways.
• The causal structure of STPs is significantly altered in tumorous tissue, offering insights into cancer mechanisms.
• These findings support the potential for identifying novel cancer driver genes and therapeutic targets.