Files
Abstract
Availability of reliable genotyping platforms for single nucleotide polymorphism markers (SNP) has made genomic breeding value (GBV) estimation a reality. Unfortunately, genotyping is still expensive and its use at large scale requires SNPs genotypes of non-typed animals to be inferred from genotyped animals. However, relationships and allele frequency information could be limited. To overcome this problem we proposed combining genotyping information from high and low density SNP panels. This low density and low cost chip will provide an additional source of information, linkage disequilibrium, in inferring missing genotypes. The proposed procedure was successful in increasing the probability of inferring true SNP genotypes for the non-typed animals by 12 to 18% depending of the simulation parameters. It increased accuracy of estimated GBVs by 3 to 12% depending on the number of SNPs and genotyped animals. These results suggest that this procedure could provide a cost effective tool for large genomic evaluation.