Improving SNP prioritization and pleiotropic architecture estimation by incorporating prior knowledge using graph-GPA

Area of Science:

• Genetics
• Bioinformatics
• Computational Biology

Background:

• Integrating multiple genome-wide association studies (GWAS) enhances the identification of genetic variants for complex traits.
• Leveraging pleiotropy, the shared genetic basis among phenotypes, can increase statistical power but presents integration challenges for large datasets.
• Previous work developed graph-GPA, a Bayesian model for integrating multiple GWAS datasets to boost power and estimate pleiotropic architecture.

Purpose of the Study:

• To propose an improved graph-GPA model incorporating external knowledge of phenotype-phenotype relationships.
• To guide the estimation of genetic correlation and association mapping using prior biological information.
• To enhance the identification of genetic variants associated with complex traits and understand their interrelationships.

Main Methods:

• Developed an enhanced graph-GPA model that integrates external knowledge on phenotype-phenotype relationships.
• Utilized a prior disease graph derived from text mining of biomedical literature.
• Applied the model to GWAS datasets for 12 complex diseases.

Main Results:

• The improved graph-GPA model demonstrated enhanced power in identifying risk genetic variants.
• The application facilitated a better understanding of the genetic relationships among complex diseases.
• External knowledge integration improved genetic correlation estimation and association mapping.

Conclusions:

• The enhanced graph-GPA model effectively integrates external phenotype relationship knowledge for improved genetic variant identification.
• This approach strengthens the analysis of pleiotropy and genetic architecture across multiple complex diseases.
• The method provides a powerful framework for dissecting the genetic underpinnings of complex traits.