Conditional and marginal SNP-heritability to leverage ancestral and environmental diversity

Area of Science:

• Genetics
• Statistical Genetics
• Population Genetics

Background:

• SNP-heritability estimates the proportion of phenotypic variance explained by SNPs in genome-wide association studies.
• Current methods often yield a single marginal heritability for diverse datasets, potentially masking subpopulation-specific heritabilities.
• Genetic and environmental exposures can differ across subpopulations, leading to varying heritabilities.

Purpose of the Study:

• To develop a conditional SNP-heritability approach for estimating multiple heritabilities within diverse datasets.
• To account for varying ancestral compositions and environmental exposures across subpopulations.
• To provide a more nuanced understanding of genetic influences on traits in heterogeneous populations.

Main Methods:

• Proposed a conditional SNP-heritability framework to estimate distinct heritabilities for different subpopulations.
• Incorporated estimation of conditional genetic and environmental variances.
• Utilized a combination of the delta method and bootstrapping for standard error calculation.

Main Results:

• Validated the conditional SNP-heritability method through extensive simulations.
• Applied the method to the Adolescent Brain Cognitive Development study dataset (6603 subjects, ages 9-11).
• Demonstrated that SNP-heritability of intelligence scores varies with socio-economic status, linked to differing extrinsic variances.

Conclusions:

• The conditional SNP-heritability approach offers a valuable tool for analyzing genetic risk differences across subpopulations.
• This method leverages data heterogeneity to uncover novel insights into trait heritability.
• Improved methodology enhances our ability to understand genetic influences in diverse populations.