Imputation Performance in Latin American Populations: Improving Rare Variants Representation With the Inclusion of Native American Genomes

• DOI: 10.3389/fgene.2021.719791
• Publication Date: 2022-01-03T08:26:07Z
• Authors: Andrés Jiménez-Ka...; Amanda Y. Chong; Adrián Cortés; Consuelo D. Quint...; Selene L. Fernand...; Leticia Ferreyra-...; Luis Pablo Cruz-H...; Santiago G. Medin...; Mashaal Sohail; María J. Palma-Ma...; Gudalupe Delgado-...; Norma Mongua-Rodr...; Alexander J. Mentzer; Adrian V. S. Hill; Hortensia Moreno-...; Alicia Huerta-Cha...; Carlos A. Aguilar...; Michael Torres; Hie Lim Kim; Namrata Kalsi; Stephan C. Schuster; Teresa Tusié-Luna; Diego Ortega Del-...; Lourdes García-Ga...; Andrés Moreno-Est...
• Keywords: Imputation (statistics); 1000 Genomes Project; Genome-wide Association Study; Genetic Association

Current Genome-Wide Association Studies (GWAS) rely on genotype imputation to increase statistical power, improve fine-mapping of association signals, and facilitate meta-analyses. Due the complex demographic history Latin America lack balanced representation Native American genomes in current panels, discovery locally relevant disease variants is likely be missed, limiting scope impact biomedical research these populations. Therefore, necessity better diversity genomic databases a scientific imperative. Here, we expand 1,000 Genomes reference panel (1KGP) with 134 (1KGP + NAT) assess performance individuals mixed ancestry. Our increased number SNPs above GWAS quality threshold, thus improving power for studies region. It also accuracy, particularly low-frequency segregating ancestry tracts. The improvement subtle but consistent across countries proportional added from local source To project potential higher genomes, performed simulations found that at least 3,000 are needed equal European This reflects concerning imbalance references highlights contribution our work reducing it while complementing efforts global equity research.