Detection of candidate tumor driver genes using a fully integrated Bayesian approach

Area of Science:

• Genomics
• Cancer Biology
• Computational Biology

Background:

• DNA copy number alterations (CNAs) significantly impact gene expression, driving disease development, particularly cancer.
• Candidate tumor driver genes, affected by CNAs, can alter downstream gene expression, promoting cancer progression.
• Identifying these driver genes is crucial for discovering novel therapeutic targets in personalized cancer treatment.

Purpose of the Study:

• To propose a novel Bayesian approach for identifying candidate tumor driver genes.
• To model DNA copy number and gene expression data jointly, capturing their dependency.
• To improve the detection of candidate tumor driver genes and understand underlying biological mechanisms.

Main Methods:

• Developed a Bayesian joint modeling approach for copy number and gene expression data.
• Modeled the dependency between copy number alterations and gene expression through conditional probabilities.
• Simultaneously identified CNAs and differentially expressed genes.

Main Results:

• The proposed joint modeling approach demonstrated improved performance in identifying candidate tumor driver genes compared to existing methods.
• The method successfully identified CNAs and differentially expressed genes simultaneously.
• Application to a head and neck squamous cell carcinoma dataset validated the approach's effectiveness.

Conclusions:

• The Bayesian joint modeling approach offers a significant advancement in identifying candidate tumor driver genes.
• This method enhances the understanding of biological processes underlying cancer development.
• The findings support the development of novel therapeutic strategies for personalized cancer therapy.