Modeling missing pedigree in single-step genomic BLUP

Abstract

The objective was to compare methods of modeling missing pedigree in single-step genomic BLUP (ssGB-LUP). Options for modeling missing pedigree included ignoring the missing pedigree, unknown parent groups (UPG) based on A (the numerator relationship matrix) or H (the unified pedigree and genomic relationship matrix), and metafounders. The assumptions for the distribution of estimated breeding values changed with the different models. We simulated data with heritabilities of 0.3 and 0.1 for dairy cattle populations that had more missing pedigrees for animals of lesser genetic merit. Predictions for the youngest generation and UPG solutions were compared with the true values for validation. For both traits, ssGBLUP with metafounders provided accurate and unbiased predictions for young animals while also appropriately accounting for genetic trend. Accuracy was least and bias was greatest for ss-GBLUP with UPG for H for the trait with heritability of 0.3 and with UPG for A for the trait with heritability of 0.1. For the trait with heritability of 0.1 and UPG for H, the UPG accuracy (SD) was −0.49 (0.12), suggesting poor estimates of genetic trend despite having little bias for validations on young, genotyped animals. Problems with UPG estimates were likely caused by the lesser amount of information available for the lower heritability trait. Hence, UPG need to be defined differently based on the trait and amount of information. More research is needed to investigate accounting for UPG in A 22 to better account for missing pedigrees for genotyped animals.

Bias and accuracy of dairy sheep evaluations using BLUP and SSGBLUP with metafounders and unknown parent groups

“…However, SSGBLUP-EUPG model was biased ( ), which indicates inflation of EBV. This agrees with [ 18 ] who found that SSGBLUP-EUPG was biased. In Fig.…”

“…Estimates of bias, dispersion and accuracy were obtained with evaluation models that used only pedigree or a combination of pedigree and genomic relationship matrices with different strategies to model missing pedigree and using the LR method. The properties of such types of models have recently been extensively investigated [ 18 , 30 , 36 – 40 ]. The current study adds further evidence that the metafounder approach should be the preferred one for genomic evaluation across species.…”

“…Whereas in “regular” SSGBLUP the only changes concern genotyped animals, here there are extensive changes that make programming difficult. In addition, because accounts correctly for the different origins and does not need pedigree completion, there is, depending on the pedigree structure, some sort of double-counting as observed by [ 18 ]. These problems are solved by MF which proposes: where is built with allele frequencies of 0.5 and there is no extra scaling to match , although there is blending as described below.…”

“…Two strategies can be used to model the missing pedigree: unknown parent groups (UPG) [ 15 , 16 ] and metafounders (MF) [ 17 ]. There is some evidence that the use of MF improves the performance of genetic evaluation [ 18 ], but it has not been systematically studied.…”

Bias in genomic predictions by mating practices for linear type traits in a large-scale genomic evaluation

“…In those years, young sons of strongly selected top proven sires, as well as old elite bulls born before 2008, were genotyped. Another reason was an improper definition of unknown parent groups (UPG) assigned for missing parents (Bradford et al, 2019;Tsuruta et al, 2019). When Bias in genomic predictions by mating practices for linear type traits in a large-scale genomic evaluation S. Tsuruta, 1 * T. J. Lawlor, 2 D. A. L. Lourenco, 1 and I. Misztal 1 animals with no phenotypes or few animals are assigned to UPG, the UPG cannot be accurately estimated.…”