Explicating heterogeneity of complex traits has strong potential for improving GWAS efficiency

Area of Science:

• Genetics
• Statistical genetics
• Human genetics

Background:

• Genome-wide association studies (GWAS) commonly employ a sample-size-centered strategy to identify genetic associations with complex traits.
• Concerns exist that this traditional approach may be reaching its limits, particularly for traits like body mass index (BMI) and lipids, due to diminishing returns in statistical power.
• Heterogeneity in genetic associations across diverse populations is increasingly recognized as a significant factor influencing GWAS outcomes.

Purpose of the Study:

• To evaluate the efficiency of the traditional sample-size-centered strategy in GWAS for complex traits.
• To investigate the impact of increasing sample size on the statistical significance of genetic associations.
• To explore the potential for improving GWAS by addressing heterogeneity, specifically biodemographic factors.

Main Methods:

• Analysis of results from the four largest GWAS meta-analyses for body mass index (BMI) and lipids.
• Examination of the relationship between sample size and p-values for genetic effects.
• Calculation of GWAS efficiency as the ratio of log-transformed p-value to sample size.
• Assessment of the prevalence and impact of heterogeneity across identified loci.

Main Results:

• Increasing sample size in large GWAS (N > 100,000) did not consistently yield smaller p-values for genetic effects; in some cases, p-values became larger.
• The efficiency of GWAS, defined by the ratio of log-transformed p-value to sample size, was higher in larger samples for a minority of loci.
• Heterogeneity in genetic associations was substantial, affecting 11-79% of loci in the analyzed GWAS, with biodemographic processes identified as a key underexplored source.

Conclusions:

• The traditional strategy of solely increasing sample size in GWAS may not be the most efficient approach for complex traits.
• Heterogeneity, particularly stemming from biodemographic factors, plays a critical role in genetic associations and represents a significant opportunity for improving GWAS.
• Future GWAS research should focus on explicating heterogeneity to enhance the discovery of genetic variants relevant to health care.