Genotype-imputation accuracy across worldwide human populations.

Area of Science:

• Genetics and Genomics
• Population Genetics
• Bioinformatics

Background:

• Genome-wide association (GWA) studies identify genetic variants associated with complex diseases.
• Imputation of ungenotyped single-nucleotide polymorphism (SNP) loci using reference panels enhances GWA study power.
• The effectiveness of imputation depends on the similarity between study samples and reference panels.

Purpose of the Study:

• To evaluate the portability and accuracy of HapMap reference panels for genotype imputation across diverse global populations.
• To determine optimal reference panel compositions for maximizing imputation accuracy in various ethnic groups.
• To provide guidance for selecting reference panels in imputation-based GWA studies of diverse populations.

Main Methods:

• Genotyped 513,008 autosomal SNP loci in 443 individuals from 29 worldwide populations.
• Assessed imputation accuracy using single HapMap panels and optimal mixtures of panels for masked genotypes.
• Evaluated imputation accuracy using SNP chip data and combinations of HapMap panels when all genotypes were unobserved.

Main Results:

• Imputation accuracy varied significantly across populations, with European populations showing the highest accuracy.
• Optimal imputation accuracy was achieved by combining individuals from at least two HapMap reference panels for most populations.
• Combinations of HapMap panels improved imputation accuracy regardless of whether genotypes were partially or fully unobserved.

Conclusions:

• The portability of HapMap reference panels for imputation is population-dependent.
• Using mixed reference panels, particularly combinations of two or more, is crucial for accurate imputation in diverse populations.
• These findings offer a practical guide for optimizing imputation strategies in GWA studies across global populations.