A population-specific reference panel empowers genetic studies of Anabaptist populations

Genotype imputation is a powerful strategy for achieving the large sample sizes required for identification of variants underlying complex phenotypes, but imputation of rare variants remains problematic. Genetically isolated populations offer one solution, however population-specific reference panel...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Scientific reports 2017-07, Vol.7 (1), p.6079-9, Article 6079
Hauptverfasser: Hou, Liping, Kember, Rachel L., Roach, Jared C., O’Connell, Jeffrey R., Craig, David W., Bucan, Maja, Scott, William K., Pericak-Vance, Margaret, Haines, Jonathan L., Crawford, Michael H., Shuldiner, Alan R., McMahon, Francis J.
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Genotype imputation is a powerful strategy for achieving the large sample sizes required for identification of variants underlying complex phenotypes, but imputation of rare variants remains problematic. Genetically isolated populations offer one solution, however population-specific reference panels are needed to assure optimal imputation accuracy and allele frequency estimation. Here we report the Anabaptist Genome Reference Panel (AGRP), the first whole-genome catalogue of variants and phased haplotypes in people of Amish and Mennonite ancestry. Based on high-depth whole-genome sequence (WGS) from 265 individuals, the AGRP contains >12 M high-confidence single nucleotide variants and short indels, of which ~12.5% are novel. These Anabaptist-specific variants were more deleterious than variants with comparable frequencies observed in the 1000 Genomes panel. About 43,000 variants showed enriched allele frequencies in AGRP, consistent with drift. When combined with the 1000 Genomes Project reference panel, the AGRP substantially improved imputation, especially for rarer variants. The AGRP is freely available to researchers through an imputation server.
ISSN:2045-2322
2045-2322
DOI:10.1038/s41598-017-05445-3