667
Accuracy of Whole-genome Sequence Genotype Imputation in Cattle Breeds

Tuesday, August 19, 2014
Posters (The Westin Bayshore)
Honghao Li , Centre for Genetic Improvement of Livestock, University of Guelph, Guelph, ON, Canada
Mehdi Sargolzaei , The Semex Alliance, Guelph, ON, Canada
Flavio S Schenkel , Centre for Genetic Improvement of Livestock, University of Guelph, Guelph, ON, Canada
Abstract Text:

Increase in accuracy of genomic predictions can possibly be achieved by using sequence genotypes. Large number of sequence genotypes can be obtained by imputation. This study compared FImpute and BEAGLE for sequence genotype imputation from Bovine Illumina HD and 54k chips using 357 sequence genotypes from six cattle breeds, including Angus, Brown Swiss, Holstein, Jersey, Limousin, and Simmental. In addition the use of a multi-breed reference population was evaluated. FImpute was more accurate and faster than BEAGLE in all scenarios investigated. As expected, imputation from HD to sequence was substantially more accurate than from 54k genotypes. Accuracy of imputation from HD to sequence genotypes was increased when a multi-breed reference population was used. FImpute imputed more accurately rare allele variants than BEAGLE. Therefore,  multi-breed reference population using FImpute is suggested for imputation from HD to sequence genotypes in the cattle breeds analyzed.

Keywords:

cattle

whole-genome sequence

genotype imputation

See more of: Posters: Methods and Tools - Genome Sequencing
See more of: Methods and Tools: Genome sequencing
<< Previous Abstract | Next Abstract >>