A Bayesian integrative genomic model for pathway analysis of complex traits.

Area of Science:

• Genomics
• Systems Biology
• Statistical Genetics

Background:

• Modern technologies generate diverse genomic data (DNA, RNA, etc.) from single samples.
• Current analyses often focus on one data type, neglecting gene-protein-reaction interactions crucial for complex traits.

Purpose of the Study:

• To develop a novel integrative model for analyzing complex phenotypes using multiple genomic data types.
• To simultaneously identify direct and indirect genomic effects on phenotypes.

Main Methods:

• A Bayesian hierarchical model combining path analysis and stochastic search variable selection.
• Simultaneous integration of multiple genomic data types.

Main Results:

• The proposed Bayesian model demonstrated increased sensitivity in detecting genomic effects in simulations.
• Application to a gemcitabine pharmacogenomic study showed improved detection capabilities.
• Outperformed standard single data type analysis in specific scenarios.

Conclusions:

• The integrative Bayesian model offers a powerful framework for leveraging multi-omics data in association studies.
• Further research is needed to optimize computational efficiency for broader application.