Genotype imputation from various low-density SNP panels and its impact on accuracy of genomic breeding values in pigs

Abstract

The uptake of genomic selection (GS) by the swine industry is still limited by the costs of genotyping. A feasible alternative to overcome this challenge is to genotype animals using an affordable low-density (LD) single nucleotide polymorphism (SNP) chip panel followed by accurate imputation to a high-density panel. Therefore, the main objective of this study was to screen incremental densities of LD panels in order to systematically identify one that balances the tradeoffs among imputation accuracy, prediction accuracy of genomic estimated breeding values (GEBVs), and genotype density (directly associated with genotyping costs). Genotypes using the Illumina Porcine60K BeadChip were available for 1378 Duroc (DU), 2361 Landrace (LA) and 3192 Yorkshire (YO) pigs. In addition, pseudo-phenotypes (de-regressed estimated breeding values) for five economically important traits were provided for the analysis. The reference population for genotyping imputation consisted of 931 DU, 1631 LA and 2103 YO animals and the remainder individuals were included in the validation population of each breed. A LD panel of 3000 evenly spaced SNPs (LD3K) yielded high imputation accuracy rates: 93.78% (DU), 97.07% (LA) and 97.00% (YO) and high correlations (>0.97) between the predicted GEBVs using the actual 60 K SNP genotypes and the imputed 60 K SNP genotypes for all traits and breeds. The imputation accuracy was influenced by the reference population size as well as the amount of parental genotype information available in the reference population. However, parental genotype information became less important when the LD panel had at least 3000 SNPs. The correlation of the GEBVs directly increased with an increase in imputation accuracy. When genotype information for both parents was available, a panel of 300 SNPs (imputed to 60 K) yielded GEBV predictions highly correlated (⩾0.90) with genomic predictions obtained based on the true 60 K panel, for all traits and breeds. For a small reference population size with no parents on reference population, it is recommended the use of a panel at least as dense as the LD3K and, when there are two parents in the reference population, a panel as small as the LD300 might be a feasible option. These findings are of great importance for the development of LD panels for swine in order to reduce genotyping costs, increase the uptake of GS and, therefore, optimize the profitability of the swine industry.

Development of a low-density panel for genomic selection of pigs in Russia1

“…In agreement with Huang et al (2012) , Cleveland and Hickey (2013) , and Wellmann et al (2013) , the presence of dense parental genotypes improved imputation accuracy, which was particularly so in the case of virtual LD panels containing very low SNP number (e.g., 300 SNP, 600 SNP). Similar to the results obtained recently by Grossi et al (2018) the minimum panel size which allowed somewhat reasonable imputation quality was the 300 SNP panel; however, sufficiently high imputation accuracy in this case could be achieved only when both parents were genotyped using HD panels. It should be noted that in the previous studies, 450 SNP ( Cleveland and Hickey, 2013 ) and 384 SNP ( Huang et al, 2012 ; Wellmann et al, 2013 ) LD panels were regarded as the most low density ones.…”

“…The 3 breeds commonly used in pork production in Russia, namely, Landrace, Duroc, and Large White were considered. The previous studies on genotype imputation from LD panels were mainly focused on the Yorkshire, Pietrain, and Landrace pig breeds (e.g., Wellmann et al, 2013 ; Badke et al, 2014 ; Xiang et al, 2015 ), while Duroc and Large White pigs have been directly considered only in a single study each ( Grossi et al, 2018 and Carillier-Jacquin, 2018 , respectively). We concentrated on these breeds in our work since they play the central role in Russian pork production, and we believe that the designed LD-chips may be of value for selection within these breeds on pig farms.…”

“…Another important factor to consider when developing LD-chips is SNP density on the LD panel. The results obtained so far by other authors have demonstrated that panels containing about 3K SNP markers might be enough to impute HD genotypes with high confidence in multiple pig breeds ( Wellmann et al, 2013 ; Grossi et al, 2018 ). Low-density panels may be also designed to include specific SNPs associated with desirable production, or reproduction traits, or disease susceptibility.…”

“…Ninety percent imputation accuracy is sufficient to obtain genomic evaluations almost identical to evaluations obtained using HD genomes ( Wellmann et al, 2013 ). Recently, several attempts to create LD-chips for use in cattle ( Zhang and Druet, 2010 ; Judge et al, 2016 ; Aliloo et al, 2018 ; Korkuć et al, 2019 ), pig ( Gualdrón Duarte et al, 2013 ; Wellmann et al, 2013 ; Xiang et al, 2015 ; Carillier-Jacquin et al, 2018 ; Grossi et al, 2018 ), as well as sheep ( Ventura et al, 2016 ; Raoul et al, 2017 ; O’ Brien et al, 2019 ), and chicken ( Wang et al, 2013 ; Herry et al, 2018 ) breeding have been published. The previous studies analyzing the prospects of using imputation from LD panels suggested different approaches for marker selection for LD-chips: random, uniform, or based on LD ( Gualdrón Duarte et al, 2013 ; Judge et al, 2016 ; Grossi et al, 2018 ).…”

“…It may also be useful to include the markers specific for a given population or associated with the traits or diseases important for local breeding programs among the markers in LD panel. It has been also pointed out that the most important factors that should be taken into account when designing an LD panel in order to increase imputation accuracy are the reference population size and the degree of kinship between the test and reference population, the rate of genotyped close relatives in the reference population, as well as the effects of minor allele frequencies (MAFs) for the imputed SNPs and LD threshold ( Cleveland and Hickey, 2013 ; Grossi et al, 2018 ; Herry et al, 2018 ). Another important factor to consider when developing LD-chips is SNP density on the LD panel.…”