graph-GPA 2.0: improving multi-disease genetic analysis with integration of functional annotation data

Area of Science:

• Genetics
• Bioinformatics
• Computational Biology

Background:

• Genome-wide association studies (GWAS) identify genetic variants linked to diseases, but functional mechanisms remain unclear, especially for shared variants across phenotypes.
• Integrating diverse data types is crucial for a comprehensive understanding of complex genetic architectures.

Purpose of the Study:

• To introduce graph-GPA 2.0 (GGPA 2.0), a novel statistical framework for integrating multiple GWAS datasets and functional annotations.
• To enhance the detection of disease-associated variants and improve the estimation of relationships among diseases.
• To elucidate the functional mechanisms underlying genetic variants, particularly those shared across different phenotypes.

Main Methods:

• GGPA 2.0 integrates GWAS data from multiple phenotypes within a unified statistical framework.
• Functional annotations from sources like GenoSkyline and GenoSkyline-Plus were incorporated.
• A prior disease graph generated by biomedical literature mining was utilized.
• Simulation studies and analysis of five autoimmune and five psychiatric disorders were performed.

Main Results:

• Incorporating functional data with GGPA 2.0 improved disease-associated variant detection and disease relationship accuracy in simulations.
• Analysis of autoimmune diseases revealed enrichment for blood-related epigenetic marks (e.g., B cells, regulatory T cells).
• Psychiatric disorders showed enrichment for brain-specific epigenetic marks (e.g., prefrontal cortex, inferior temporal lobe).
• GGPA 2.0 detected pleiotropy between bipolar disorder and schizophrenia and demonstrated robustness to irrelevant functional annotations.

Conclusions:

• GGPA 2.0 is a powerful tool for identifying phenotype-specific and shared genetic variants across multiple diseases.
• The framework aids in understanding the functional mechanisms of associated variants by integrating diverse data.
• GGPA 2.0 provides a robust approach for genetic analysis, even with imperfect functional annotation data.