A Bayesian genomic selection approach incorporating prior feature ordering and population structures with application...

Area of Science:

• Genetics
• Biostatistics
• Cardiovascular Disease Research

Background:

• Coronary artery disease (CAD) is a leading cause of death, with significant genetic influences in both sexes.
• Identifying specific genetic variants associated with CAD is crucial for understanding disease mechanisms and improving risk prediction.

Purpose of the Study:

• To propose a novel Bayesian variable selection framework for identifying genetic variants associated with coronary artery disease (CAD) status.
• To develop an innovative prior that accounts for the ordering structure of genetic variants, improving selection accuracy.
• To incorporate population structure by fitting separate regressions for different subject groups, enhancing disease risk reflection.

Main Methods:

• Developed a Bayesian variable selection framework incorporating an innovative prior for genetic variant inclusion probabilities, considering their ordering.
• Implemented a method to group subjects based on population structure and fit separate regression models.
• Utilized a Markov random field-inspired prior to borrow strength across regression models, enhancing model performance.

Main Results:

• The proposed framework demonstrated improved variable selection and prediction performance in simulation studies.
• The method was successfully applied to the CATHeterization GENetics (CG) dataset for binary CAD status.
• The novel approach effectively identifies important genetic variants and accounts for population heterogeneity.

Conclusions:

• The novel Bayesian framework offers improved accuracy in identifying genetic variants associated with coronary artery disease.
• Accounting for variant ordering and population structure enhances the precision of genetic association studies for CAD.
• This approach provides a robust tool for genetic research in cardiovascular diseases.