Bivariate GSA-MiXeR: A Novel Tool for Functional Genomic Analyses Implicates Diverse Neural Cell Types for...

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Genomic discoveries for complex human traits are increasing.
• Advanced functional genomics tools are needed to interpret polygenic and pleiotropic architectures.
• Understanding genetic overlap between traits is essential for biological insights.

Purpose of the Study:

• Introduce bivariate GSA-MiXeR, a novel tool for functional genomics analysis.
• Model partitioned heritability and covariance of two traits within a genomic region of interest (ROI).
• Estimate trait-specific fold enrichment, local genetic correlation, and local genetic omnibus statistic for ranking ROIs.

Main Methods:

• Developed bivariate GSA-MiXeR for analyzing genetic architecture of complex traits.
• Implemented modeling of heritability and covariance within genomic ROIs.
• Validated the tool through simulations and applied analyses using single-cell RNA sequencing data.

Main Results:

• Bivariate GSA-MiXeR successfully models partitioned heritability and covariance.
• The tool estimates local genetic correlations in both continuous and disjoint ROIs.
• Applied analyses implicated diverse neural cell types in psychiatric and neurodegenerative disorders.

Conclusions:

• Bivariate GSA-MiXeR expands the assessment of local genetic overlap among complex traits.
• The tool facilitates the transition from genetic discovery to mechanistic insights.
• Provides new analytical avenues for functional genomics research.