Estimating cross-population genetic correlations of causal effect sizes

Area of Science:

• Population Genetics
• Complex Trait Genetics
• Genomic Architecture

Background:

• Studies examine genetic correlations of single-nucleotide polymorphism (SNP) effect sizes across populations to understand complex trait genetic architectures.
• Cross-population correlation of joint-fit effect sizes (ρg) is influenced by true causal effect correlations (ρb) and linkage disequilibrium (LD) patterns.

Purpose of the Study:

• Derive the ratio of cross-population genetic correlations (ρg/ρb) as a function of LD.
• Develop a method to estimate the correlation of true causal effect sizes (ρb) across populations.

Main Methods:

• Applied existing methods to estimate ρg.
• Derived the ratio ρg/ρb based on population-specific LD patterns.
• Utilized the derived ratio to estimate ρb.

Main Results:

• Estimated ρb between Europeans and East Asians (n=9 traits) as 0.55 (SE=0.14).
• Estimated ρb between Europeans and South Asians (n=13 traits) as 0.54 (SE=0.18).
• Estimated ρb for type 2 diabetes and rheumatoid arthritis between Europeans and East Asians as 0.48 (SE=0.06) and 0.65 (SE=0.09), respectively.

Conclusions:

• The findings suggest substantially different causal genetic architectures across continental populations.
• The developed method allows for estimation of true genetic effect correlations, independent of LD differences.