Multiethnic genetic association studies improve power for locus discovery

Area of Science:

• Genetics
• Population Genetics
• Genomic Association Studies

Background:

• Genome-wide association studies (GWAS) have historically concentrated on European ancestry populations.
• Next-generation sequencing advances enable cataloging and testing low-frequency genetic variations for disease association.
• Current approaches may miss crucial genetic insights due to limited ancestral diversity.

Purpose of the Study:

• To evaluate the statistical power of association studies incorporating multiple ethnicities compared to European-ancestry-only studies.
• To quantify the benefits of including diverse populations for detecting genetic associations, especially for low-frequency variants.
• To assess the contribution of different ethnic groups to the power of genetic discovery.

Main Methods:

• Simulated genetic models based on empirical allele frequency distributions from diverse populations (HapMap Phase 3, 1000 Genomes Project).
• Modeled a two-phase GWAS approach: initial phase in European samples, followed by a second phase in other populations or a multiethnic design.
• Quantified power gains across various genetic models and allele frequencies.

Main Results:

• Significant power gains are achievable by including worldwide populations in whole-genome association studies.
• African populations offer the largest relative power increases for detecting low-frequency (<5%) variants of moderate effect.
• European-ancestry-focused studies exhibit limited power for certain low-frequency variants.

Conclusions:

• Broadening genetic studies to include diverse global populations is crucial for efficient discovery of genetic loci influencing phenotypic variation.
• Future genomic studies should prioritize multiethnic cohorts to maximize the detection of disease-associated variants.
• Incorporating diverse ancestries enhances the power to identify genetic contributions to traits, especially when European-ancestry datasets are saturated.