Genetic variants influencing phenotypic variance heterogeneity

Area of Science:

• Genomics
• Epigenetics
• Statistical Genetics

Background:

• Genetic studies typically focus on variants affecting trait means, overlooking those influencing trait variance.
• Variance heterogeneity, or how trait variance differs across genotypes, is less explored but crucial for understanding complex phenotypes.
• Identifying genetic variants associated with DNA methylation variance heterogeneity can reveal novel biological mechanisms or statistical artifacts.

Purpose of the Study:

• To identify variance single-nucleotide polymorphisms (vSNPs) associated with DNA methylation levels genome-wide.
• To investigate whether identified vSNPs represent true biological interactions (gene × gene, gene × environment) or statistical artifacts.
• To understand the underlying genetic architecture of DNA methylation variance heterogeneity.

Main Methods:

• Genome-wide association study (GWAS) for variance heterogeneity in DNA methylation.
• Analysis of genotype data (over 10 million SNPs) and DNA methylation data (over 430,000 CpG sites) in 729 individuals.
• Statistical analysis to differentiate biological interactions from linkage disequilibrium (LD) effects.

Main Results:

• Identified 7195 vSNPs associated with DNA methylation variance heterogeneity at specific CpG sites (P < 9.4 × 10-11).
• This number is substantially lower than the 52,335 CpG sites associated with mean DNA methylation levels.
• Variance heterogeneity was predominantly explained by additional, often rare, SNPs in high LD with the identified vSNPs, or co-segregating variants.

Conclusions:

• Variance heterogeneity in DNA methylation is mainly driven by the cumulative effects of multiple SNPs, rather than distinct biological interactions.
• These findings suggest that statistical artifacts, specifically complex SNP-SNP relationships, are primary drivers of observed variance heterogeneity.
• Understanding these multi-SNP effects is critical for accurate interpretation of variance heterogeneity in both DNA methylation and complex clinical phenotypes.