High Imputation Accuracy in Layer Chicken from Sequence Data on a Few Key Ancestors

Abstract Text: We assessed a scenario designed to mimic the imputation of full genome sequence data in White layer chickens, genotyped at medium (60K) density. Factors affecting accuracy were the size of the reference population, the level of the relationship between the reference and test populations and minor allele frequency of the SNP being imputed. Genotype imputation based on 22 or 62 carefully selected reference animals resulted in accuracies between 0.78 and 0.87. So, a very small reference population already provided satisfactory results. These results suggest that full genome SNP imputation is possible in layer chicken when a suitable pool of key ancestors is sequenced. SNPs with low MAF were more difficult to impute. Accuracies did not reduce when test populations were 1, 2, or 3 generations away from the reference animals.

Keywords:

Layer chicken

Imputation accuracy

Whole genome sequence

Key ancestors