Validation of single-step GBLUP genomic predictions from threshold models using the linear regression method: An application in chicken mortality

Use of the linear regression method to evaluate population accuracy of predictions from non-linear models

Background

To address the limitations of commonly used cross-validation methods, the linear regression method (LR) was proposed to estimate population accuracy of predictions based on the implicit assumption that the fitted model is correct. This method also provides two statistics to determine the adequacy of the fitted model. The validity and behavior of the LR method have been provided and studied for linear predictions but not for nonlinear predictions. The objectives of this study were to 1) provide a mathematical proof for the validity of the LR method when predictions are based on conditional means, regardless of whether the predictions are linear or non-linear 2) investigate the ability of the LR method to detect whether the fitted model is adequate or inadequate, and 3) provide guidelines on how to appropriately partition the data into training and validation such that the LR method can identify an inadequate model.

Results

We present a mathematical proof for the validity of the LR method to estimate population accuracy and to determine whether the fitted model is adequate or inadequate when the predictor is the conditional mean, which may be a non-linear function of the phenotype. Using three partitioning scenarios of simulated data, we show that the one of the LR statistics can detect an inadequate model only when the data are partitioned such that the values of relevant predictor variables differ between the training and validation sets. In contrast, we observed that the other LR statistic was able to detect an inadequate model for all three scenarios.

Conclusion

The LR method has been proposed to address some limitations of the traditional approach of cross-validation in genetic evaluation. In this paper, we showed that the LR method is valid when the model is adequate and the conditional mean is the predictor, even when it is a non-linear function of the phenotype. We found one of the two LR statistics is superior because it was able to detect an inadequate model for all three partitioning scenarios (i.e., between animals, by age within animals, and between animals and by age) that were studied.

Confidence intervals for validation statistics with data truncation in genomic prediction

BACKGROUND

Validation by data truncation is a common practice in genetic evaluations because of the interest in predicting the genetic merit of a set of young selection candidates. Two of the most used validation methods in genetic evaluations use a single data partition: predictivity or predictive ability (correlation between pre-adjusted phenotypes and estimated breeding values (EBV) divided by the square root of the heritability) and the linear regression (LR) method (comparison of "early" and "late" EBV). Both methods compare predictions with the whole dataset and a partial dataset that is obtained by removing the information related to a set of validation individuals. EBV obtained with the partial dataset are compared against adjusted phenotypes for the predictivity or EBV obtained with the whole dataset in the LR method. Confidence intervals for predictivity and the LR method can be obtained by replicating the validation for different samples (or folds), or bootstrapping. Analytical confidence intervals would be beneficial to avoid running several validations and to test the quality of the bootstrap intervals. However, analytical confidence intervals are unavailable for predictivity and the LR method.

RESULTS

We derived standard errors and Wald confidence intervals for the predictivity and statistics included in the LR method (bias, dispersion, ratio of accuracies, and reliability). The confidence intervals for the bias, dispersion, and reliability depend on the relationships and prediction error variances and covariances across the individuals in the validation set. We developed approximations for large datasets that only need the reliabilities of the individuals in the validation set. The confidence intervals for the ratio of accuracies and predictivity were obtained through the Fisher transformation. We show the adequacy of both the analytical and approximated analytical confidence intervals and compare them versus bootstrap confidence intervals using two simulated examples. The analytical confidence intervals were closer to the simulated ones for both examples. Bootstrap confidence intervals tend to be narrower than the simulated ones. The approximated analytical confidence intervals were similar to those obtained by bootstrapping.

CONCLUSIONS

Estimating the sampling variation of predictivity and the statistics in the LR method without replication or bootstrap is possible for any dataset with the formulas presented in this study.

Comparison of different validation methods for single-step genomic evaluations based on a simulated cattle population

The validation of estimated breeding values from single-step genomic BLUP (ssGBLUP) is an important topic, as more and more countries and animal populations are currently changing their genomic prediction to single-step. The objective of this work was to compare different methods to validate single-step genomic breeding values (GEBV). The investigations were carried out using a simulation study based on the German-Austrian-Czech Fleckvieh population. To test the validation methods under different conditions, several biased and unbiased scenarios were simulated. The application of the widely used Interbull GEBV test to the single-step method is only possible to a limited extent, partly because of genomic preselection, which biases conventional estimated breeding values. Alternative validation methods considered in the study are the linear regression method proposed by Legarra and Reverter, the improved genomic validation including additional regressions as suggested by VanRaden and an adaptation of the Interbull GEBV test using daughter yield deviations (DYD) from ssGBLUP instead of pedigree BLUP. The comparison of the different methods for the different scenarios showed that for males the methods based on GEBV estimate the dispersion more accurate and less biased compared with the GEBV test using DYD from ssGBLUP, whereas the standard Interbull GEBV test is highly affected by genomic preselection for males. For females, the GEBV test using yield deviations from ssGBLUP results in better estimations for the true dispersion.

Validating genomic prediction for nitrogen efficiency index and its composition traits of Holstein cows in early lactation

Nitrogen (N) use efficiency (NUE) is an economically important trait for dairy cows. Recently, we proposed a new N efficiency index (NEI), that simultaneously considers both NUE and N pollution. This study aimed to validate the genomic prediction for NEI and its composition traits and investigate the relationship between SNP effects estimated directly from NEI and indirectly from its composition traits. The NEI composition included genomic estimated breeding value of N intake (NINT), milk true protein N (MTPN) and milk urea N yield. The edited data were 132,899 records on 52,064 cows distributed in 773 herds. The pedigree contained 122,368 animals. Genotypic data of 566,294 SNP was available for 4514 individuals. A total of 4413 cows (including 181 genotyped) and 56 bulls (including 32 genotyped) were selected as the validation populations. The linear regression method was used to validate the genomic prediction of NEI and its composition traits using best linear unbiased prediction (BLUP) and single-step genomic BLUP (ssGBLUP). The mean theoretical accuracies of validation populations obtained from ssGBLUP were higher than those obtained from BLUP for both NEI and its composition traits, ranging from 0.57 (MTPN) to 0.72 (NINT). The highest mean prediction accuracies for NEI and its composition traits were observed for the genotyped cows estimated under ssGBLUP, ranging from 0.48 (MTPN) to 0.66 (NINT). Furthermore, the SNP effects estimated from NEI composition traits, multiplied by the relative weight were the same as those estimated directly from NEI. This study preliminary showed that genomic prediction can be used for NEI, however, we acknowledge the need for further validation of this result in a larger dataset. Moreover, the SNP effects of NEI can be indirectly calculated using the SNP effects estimated from its composition traits. This study provided a basis for adding genomic information to establish NEI as part of future routine genomic evaluation programs.

Prediction ability of an alternative multi-trait genomic evaluation for residual feed intake

Selection for feed efficiency is the goal for many genetic breeding programs in beef cattle. Residual feed intake has been included in genetic evaluations to reduce feed intake without compromising performance traits as liveweight, body gain or carcass traits. However, measuring feed intake is expensive, and only a small percentage of selection candidates are phenotyped. Genomic selection has become a very important tool to achieve effective genetic progress in these traits. Another effective strategy has been the implementation of multi-trait prediction using easily recordable predictor traits on both reference animals and candidates without phenotypes, and this could be another inexpensive way to increase accuracy. The objective of this work was to analyse and compare the prediction ability of two alternative different approaches to predict GEBVs for RFI. The population of inference was Hereford bulls in Uruguay that were genotyped candidates for to selection. The first model was the conventional univariate model for RFI and the second model was a multi-trait model which included a predictor trait (weaning weight, WW), in addition to the traits used in the first one (dry matter intake, metabolic mid test weight, average daily gain and ultrasound back fat) (DMI, MWT, ADG, UBF, respectively). GEBVs from the multi-trait model were combined using selection index theory to derive RFI values. All analyses were performed using ssGBLUP procedure. The prediction ability of both models was tested using two validation strategies (30 different replicates of random groups of animals and validation across 9 different feed intake tests). The prediction quality was assessed by the following parameters: bias, dispersion, ratio of accuracies and the relative increase in accuracy by adding phenotypic information. All parameters showed that the univariate model outperforms the multi-trait model, regardless of the validation strategy considered. These results indicate that including WW as a proxy trait in a multi-trait analysis does not improve the prediction ability when all animals to be predicted are genotyped.

Genomic prediction of fertility and calving traits in Holstein cattle based on models including epistatic genetic effects

The goal of this study was to investigate whether the inclusion of genomic information and epistatic (additive by additive) genetic effects would increase the accuracy of predicting phenotypes adjusted for known environmental effects, reduce prediction bias and minimize the confounding between additive and additive by additive epistatic effects on fertility and calving traits in Holstein cattle. Phenotypic and genotypic records were available for 6090 cows. Eight cow traits were assessed including 56-day nonreturn rate (NRR), number of services (NS), calving to first insemination (CTFS), first insemination to conception (FSTC), gestation length (GL), calving ease (CE), stillbirth (SB) and calf size (CZ). Four scenarios were assessed for their ability to predict adjusted phenotypes, which included: (1) traditional pedigree-based Best Linear Unbiased Prediction (P-BLUP) for additive genetic effects (PA); (2) P-BLUP for additive and epistatic (additive by additive) genetic effects (PAE); (3) genomic BLUP (G-BLUP) for additive genetic effects (GA); and (4) G-BLUP for additive and epistatic genetic effects (GAEn, where n = 1-3 depending on the alternative ways to construct the epistatic genomic matrix used). Constructing epistatic relationship matrix as the Hadamard product of the additive genomic relationship matrix (GAE1), which is the usual method and implicitly assumes a model that fits all pairwise interactions between markers twice and includes the interactions of the markers with themselves (dominance). Two additional constructions of the epistatic genomic relationship matrix were compared to test whether removing the double counting of interactions and the interaction of the markers with themselves (GAE2), and removing double counting of interactions between markers, but including the interaction of the markers with themselves (GAE3) would had an impact on the prediction and estimation error correlation (i.e. confounding) between additive and epistatic genetic effects. Fitting epistatic genetic effects explained up to 5.7% of the variance for NRR (GAE3), 7.7% for NS (GAE1), 11.9% for CTFS (GAE3), 11.1% for FSTC (GAE2), 25.7% for GL (GAE1), 2.3% for CE (GAE1), 14.3% for SB (GAE3) and 15.2% for CZ (GAE1). Despite a substantial proportion of variance being explained by epistatic effects for some traits, the prediction accuracies were similar or lower for GAE models compared with pedigree models and genomic models without epistatic effects. Although the prediction accuracy of direct genomic values did not change significantly between the three variations of the epistatic genetic relationship matrix used, removing the interaction of the markers with themselves reduced the confounding between additive and additive by additive epistatic effects. These results suggest that epistatic genetic effects contribute to the variance of some fertility and calving traits in Holstein cattle. However, the inclusion of epistatic genetic effects in the genomic prediction of these traits is complex and warrant further investigation.

Including genotypic information in genetic evaluations increases the accuracy of sheep breeding values

The impact of inclusion of genome-wide genotypes into breeding value predictions for UK Texel sheep is addressed in this article. The main aim was to investigate the level of change in the accuracy values for EBVs when information from animal genotypes is incorporated into the genetic evaluations. New genetic parameters for a range of lamb growth, carcass composition and health traits are described and applied in the estimation of conventional breeding values (EBVs) for almost 822,000 animals as well as genomic breeding values (gEBVs) after adding 10,143 genotypes. Principal component analyses showed that there are no major distinct groups; hence, the population is mainly homogenous and genetically well-linked. Results suggested that the highest change in accuracy was observed for the animals that are not phenotyped but have good links to the reference population. This was seen especially for the lowly heritable health traits thereby proving that the use of genotypes in breeding values estimation may accelerate the genetic gain by producing more accurate values especially for young, un-phenotyped animals.

Genomic evaluation for two-way crossbred performance in cattle

BACKGROUND

Dairy cattle production systems are mostly based on purebreds, but recently the use of crossbreeding has received increased interest. For genetic evaluations including crossbreds, several methods based on single-step genomic best linear unbiased prediction (ssGBLUP) have been proposed, including metafounder ssGBLUP (MF-ssGBLUP) and breed-specific ssGBLUP (BS-ssGBLUP). Ideally, models that account for breed effects should perform better than simple models, but knowledge on the performance of these methods is lacking for two-way crossbred cattle. In addition, the differences in the estimates of genetic parameters (such as the genetic variance component and heritability) between these methods have rarely been investigated. Therefore, the aims of this study were to (1) compare the estimates of genetic parameters for average daily gain (ADG) and feed conversion ratio (FCR) between these methods; and (2) evaluate the impact of these methods on the predictive ability for crossbred performance.

METHODS

Bivariate models using standard ssGBLUP, MF-ssGBLUP and BS-ssGBLUP for the genetic evaluation of ADG and FCR were investigated. To measure the predictive ability of these three methods, we estimated four estimators, bias, dispersion, population accuracy and ratio of population accuracies, using the linear regression (LR) method.

RESULTS

The results show that, for both ADG and FCR, the heritabilities were low with the three methods. For FCR, the differences in the estimated genetic parameters were small between the three methods, while for ADG, those estimated with BS-ssGBLUP deviated largely from those estimated with the other two methods. Bias and dispersion were similar across the three methods. Population accuracies for both ADG and FCR were always higher with MF-ssGBLUP than with ssGBLUP, while with BS-ssGBLUP the population accuracy was highest for FCR and lowest for ADG.

CONCLUSIONS

Our results indicate that in the genetic evaluation for crossbred performance in a two-way crossbred cattle production system, the predictive ability of MF-ssGBLUP and BS-ssGBLUP is greater than that of ssGBLUP, when the estimated variance components are consistent across the three methods. Compared with BS-ssGBLUP, MF-ssGBLUP is more robust in its superiority over ssGBLUP.

Genomic evaluation of commercial herds with different pedigree structures using the single-step genomic BLUP in Nelore cattle

The aim of this work was to evaluate the impact of applying genomic information in pedigree uncertainty situations on genetic evaluations for growth- and cow productivity-related traits in Nelore commercial herds. Records for accumulated cow productivity (ACP) and adjusted weight at 450 days of age (W450) were used, as well as genotypes of registered and commercial herd animals, genotyped with the Clarifide Nelore 3.1 panel (~29,000 SNPs). The genetic values for commercial and registered populations were estimated using different approaches that included (ssGBLUP) or did not include genomic information (BLUP), with different pedigree structures. Different scenarios were tested, varying the proportion of young animals with unknown sires (0, 25, 50, 75, and 100%), and unknown maternal grandsires (0, 25, 50, 75, and 100%). The prediction accuracies and abilities were calculated. The estimated breeding value accuracies decreased as the proportion of unknown sires and maternal grandsires increased. The genomic estimated breeding value accuracy using the ssGBLUP was higher in scenarios with a lower proportion of known pedigree when compared to the BLUP methodology. The results obtained with the ssGBLUP showed that it is possible to obtain reliable direct and indirect predictions for young animals from commercial herds without pedigree structure.

Genomic selection models substantially improve the accuracy of genetic merit predictions for fillet yield and body weight in rainbow trout using a multi-trait model and multi-generation progeny testing

BACKGROUND

In aquaculture, the proportion of edible meat (FY = fillet yield) is of major economic importance, and breeding animals of superior genetic merit for this trait can improve efficiency and profitability. Achieving genetic gains for fillet yield is possible using a pedigree-based best linear unbiased prediction (PBLUP) model with direct and indirect selection. To investigate the feasibility of using genomic selection (GS) to improve FY and body weight (BW) in rainbow trout, the prediction accuracy of GS models was compared to that of PBLUP. In addition, a genome-wide association study (GWAS) was conducted to identify quantitative trait loci (QTL) for the traits. All analyses were performed using a two-trait model with FY and BW, and variance components, heritability, and genetic correlations were estimated without genomic information. The data used included 14,165 fish in the pedigree, of which 2742 and 12,890 had FY and BW phenotypic records, respectively, and 2484 had genotypes from the 57K single nucleotide polymorphism (SNP) array.

RESULTS

The heritabilities were moderate, at 0.41 and 0.33 for FY and BW, respectively. Both traits were lowly but positively correlated (genetic correlation; r = 0.24), which suggests potential favourable correlated genetic gains. GS models increased prediction accuracy compared to PBLUP by up to 50% for FY and 44% for BW. Evaluations were found to be biased when validation was performed on future performances but not when it was performed on future genomic estimated breeding values.

CONCLUSIONS

The low but positive genetic correlation between fillet yield and body weight indicates that some improvement in fillet yield may be achieved through indirect selection for body weight. Genomic information increases the prediction accuracy of breeding values and is an important tool to accelerate genetic progress for fillet yield and growth in the current rainbow trout population. No significant QTL were found for either trait, indicating that both traits are polygenic, and that marker-assisted selection will not be helpful to improve these traits in this population.

Removing data and using metafounders alleviates biases for all traits in Lacaune dairy sheep predictions

Bias in dairy genetic evaluations, when it exists, has to be understood and properly addressed. The origin of biases is not always clear. We analyzed 40 yr of records from the Lacaune dairy sheep breeding program to evaluate the extent of bias, assess possible corrections, and emit hypotheses on its origin. The data set included 7 traits (milk yield, fat and protein contents, somatic cell score, teat angle, udder cleft, and udder depth) with records from 600,000 to 5 million depending on the trait, ∼1,900,000 animals, and ∼5,900 genotyped elite artificial insemination rams. For the ∼8% animals with missing sire, we fit 25 unknown parent groups. We used the linear regression method to compare "partial" and "whole" predictions of young rams before and after progeny testing, with 7 cut-off points, and we obtained estimates of their bias, (over)dispersion, and accuracy in early proofs. We tried (1) several scenarios as follows: multiple or single trait, the "official" (routine) evaluation, which is a mixture of both single and multiple trait, and "deletion" of data before 1990; and (2) several models as follows: BLUP and single-step genomic (SSG)BLUP with fixed unknown parent groups or metafounders, where, for metafounders, their relationship matrix gamma was estimated using either a model for inbreeding trend, or base allele frequencies estimated by peeling. The estimate of gamma obtained by modeling the inbreeding trend resulted in an estimated increase of inbreeding, based on markers, faster than the pedigree-based one. The estimated genetic trends were similar for most models and scenarios across all traits, but were shrunken when gamma was estimated by peeling. This was due to shrinking of the estimates of metafounders in the latter case. Across scenarios, all traits showed bias, generally as an overestimate of genetic trend for milk yield and an underestimate for the other traits. As for the slope, it showed overdispersion of estimated breeding values for all traits. Using multiple-trait models slightly reduced the overestimate of genetic trend and the overdispersion, as did including genomic information (i.e., SSGBLUP) when the gamma matrix was estimated by the model for inbreeding trend. However, only deletion of historical data before 1990 resulted in elimination of both kind of biases. The SSGBLUP resulted in more accurate early proofs than BLUP for all traits. We considered that a snowball effect of small errors in each genetic evaluation, combined with selection, may have resulted in biased evaluations. Improving statistical methods reduced some bias but not all, and a simple solution for this data set was to remove historical records.

Inclusion of Sire by Herd interaction effect in the genomic evaluation for weaning weight of American Angus

A spurious negative genetic correlation between direct and maternal effects of weaning weight (WW) in beef cattle has historically been problematic for researchers and industry. Previous research has suggested the covariance between sires and herds may be contributing to this relationship. The objective of this study was to estimate the variance components (VC) for WW in American Angus with and without sire by herd (S×H) interaction effect when genomic information is used or not. Five subsets of ~100k animals for each subset were used. When genomic information was included, genotypes were added for 15,637 animals. Five replicates were performed. Four different models were tested, namely, M1: without S×H interaction effect and with covariance between direct and maternal effect (σam) ≠ 0; M2: with S×H interaction effect and σam ≠ 0; M3: without S×H interaction effect and with σam = 0; M4: with S×H interaction effect and σam = 0. VC were estimated using the restricted maximum likelihood (REML) and single-step genomic REML (ssGREML) with the average information algorithm. Breeding values were computed using single-step genomic BLUP for the models above and one additional model, which had the covariance zeroed after the estimation of VC (M5). The ability of each model to predict future breeding values was investigated with the linear regression method. Under REML, when the S×H interaction effect was added to the model, both direct and maternal genetic variances were greatly reduced, and the negative covariance became positive (i.e., when moving from M1 to M2). Similar patterns were observed under ssGREML, but with less reduction in the direct and maternal genetic variances and still a negative covariance. Models with the S×H interaction effect (M2 and M4) had a better fit according to the Akaike information criteria. Breeding values from those models were more accurate and had less bias than the other three models. The rankings and breeding values of artificial insemination sires (N = 1,977) greatly changed when the S×H interaction effect was fit in the model. Although the S×H interaction effect accounted for 3% to 5% of the total phenotypic variance and improved the model fit, this change in the evaluation model will cause severe reranking among animals.

A spurious negative genetic correlation between direct and maternal effects of weaning weight (WW) in beef cattle has been problematic for researchers and industry. Previous research suggested the covariance between sires and herds may contribute to this relationship. The objective of this study was to estimate the variance components (VC) for WW in American Angus with and without sire by herd (S×H) interaction effect when genomic information is used or not. Four models were designed to investigate the S×H effect. The restricted maximum likelihood (REML) and single-step genomic REML (ssGREML) were used to estimate VC. Breeding values were computed using single-step genomic BLUP and the validation was done through the linear regression method. Under REML, when the S×H was added to the model, both direct and maternal genetic variances were greatly reduced, and the negative covariance became positive. Similar patterns were observed under ssGREML, but with less reduction in the direct and maternal genetic variances and still a negative covariance. Breeding values from models with S×H were more accurate and had less bias than the other models. Although the S×H improved the model, this change in the evaluation model will cause severe reranking among key animals.

Determining Heat Stress Effects of Multiple Genetic Traits in Tropical Dairy Cattle Using Single-Step Genomic BLUP

Heat stress is becoming a significant problem in dairy farming, especially in tropical countries, making accurate genetic selection for heat tolerance a priority. This study investigated the effect of heat stress manifestation on genetics for milk yield, milk quality, and dairy health traits with and without genomic information using single-step genomic best linear unbiased prediction (ssGBLUP) and BLUP in Thai−Holstein crossbred cows. The dataset contained 104,150 test-day records from the first lactation of 15,380 Thai−Holstein crossbred cows. A multiple-trait random regression test-day model on a temperature−humidity index (THI) function was used to estimate the genetic parameters and genetic values. Heat stress started at a THI of 76, and the heritability estimates ranged from moderate to low. The genetic correlation between those traits and heat stress in both BLUP methods was negative. The accuracy of genomic predictions in the ssGBLUP method was higher than the BLUP method. In conclusion, heat stress negatively impacted milk production, increased the somatic cell score, and disrupted the energy balance. Therefore, in dairy cattle genetic improvement programs, heat tolerance is an important trait. The new genetic evaluation method (ssGBLUP) should replace the traditional method (BLUP) for more accurate genetic selection.

Development of genomic predictions for Angus cattle in Brazil incorporating genotypes from related American sires

Genomic prediction has become the new standard for genetic improvement programs, and currently, there is a desire to implement this technology for the evaluation of Angus cattle in Brazil. Thus, the main objective of this study was to assess the feasibility of evaluating young Brazilian Angus (BA) bulls and heifers for 12 routinely recorded traits using single-step genomic BLUP (ssGBLUP) with and without genotypes from American Angus (AA) sires. The second objective was to obtain estimates of effective population size (N_e) and linkage disequilibrium (LD) in the Brazilian Angus population. The dataset contained phenotypic information for up to 277,661 animals belonging to the Promebo breeding program, pedigree for 362,900, of which 1,386 were genotyped for 50k, 77k, and 150k single nucleotide polymorphism (SNP) panels. After imputation and quality control, 61,666 SNPs were available for the analyses. In addition, genotypes from 332 American Angus (AA) sires widely used in Brazil were retrieved from the AA Association database to be used for genomic predictions. Bivariate animal models were used to estimate variance components, traditional EBV, and genomic EBV (GEBV). Validation was carried out with the linear regression method (LR) using young-genotyped animals born between 2013 and 2015 without phenotypes in the reduced dataset and with records in the complete dataset. Validation animals were further split into progeny of BA and AA sires to evaluate if their progenies would benefit by including genotypes from AA sires. The N_e was 254 based on pedigree and 197 based on LD, and the average LD (±SD) and distance between adjacent single nucleotide polymorphisms (SNPs) across all chromosomes were 0.27 (±0.27) and 40743.68 bp, respectively. Prediction accuracies with ssGBLUP outperformed BLUP for all traits, improving accuracies by, on average, 16% for BA young bulls and heifers. The GEBV prediction accuracies ranged from 0.37 (total maternal for weaning weight and tick count) to 0.54 (yearling precocity) across all traits, and dispersion (LR coefficients) fluctuated between 0.92 and 1.06. Inclusion of genotyped sires from the AA improved GEBV accuracies by 2%, on average, compared to using only the BA reference population. Our study indicated that genomic information could help us to improve GEBV accuracies and hence genetic progress in the Brazilian Angus population. The inclusion of genotypes from American Angus sires heavily used in Brazil just marginally increased the GEBV accuracies for selection candidates.

There was a desire to implement genomic selection for Angus cattle in Brazil since the technology has been proved to increase genetic gain in animal breeding programs. Single-step genomic best linear unbiased prediction (ssGBLUP), which simultaneously combines pedigree and genomic information, was used to estimate individuals’ genomic breeding values (GEBV) or genetic merit. Genomic selection can accelerate genetic progress by increasing accuracy, especially in young animals without progeny. The accuracy of GEBV can also be improved by combing data from other countries to increase the reference population (i.e., genotyped and phenotyped animals) in small, genotyped populations. Thus, the main objective of this study was to evaluate the accuracy of GEBV for young Brazilian Angus (BA) bulls and heifers with ssGBLUP, including or not the genotypes from American Angus sires. The accuracies with ssGBLUP were higher than those from traditional BLUP (EBV calculated from pedigree), improving accuracies by, on average, 16% for young bulls and heifers. Including genotypes from American Angus sires heavily used in Brazil just marginally increased the GEBV accuracies for selection candidates.

Validation of single-step genomic predictions using the linear regression method for milk yield and heat tolerance in a Thai-Holstein population

Background and Aim:

Genomic selection improves accuracy and decreases the generation interval, increasing the selection response. This study was conducted to assess the benefits of using single-step genomic best linear unbiased prediction (ssGBLUP) for genomic evaluations of milk yield and heat tolerance in Thai-Holstein cows and to test the value of old phenotypic data to maintain the accuracy of predictions.

Materials and Methods:

The dataset included 104,150 milk yield records collected from 1999 to 2018 from 15,380 cows. The pedigree contained 33,799 animals born between 1944 and 2016, of which 882 were genotyped. Analyses were performed with and without genomic information using ssGBLUP and BLUP, respectively. Statistics for bias, dispersion, the ratio of accuracies, and the accuracy of estimated breeding values were calculated using the linear regression (LR) method. A partial dataset excluded the phenotypes of the last generation, and 66 bulls were identified as validation individuals.

Results:

Bias was considerable for BLUP (0.44) but negligible (−0.04) for ssGBLUP; dispersion was similar for both techniques (0.84 vs. 1.06 for BLUP and ssGBLUP, respectively). The ratio of accuracies was 0.33 for BLUP and 0.97 for ssGBLUP, indicating more stable predictions for ssGBLUP. The accuracy of predictions was 0.18 for BLUP and 0.36 for ssGBLUP. Excluding the first 10 years of phenotypic data (i.e., 1999-2008) decreased the accuracy to 0.09 for BLUP and 0.32 for ssGBLUP. Genomic information doubled the accuracy and increased the persistence of genomic estimated breeding values when old phenotypes were removed.

Conclusion:

The LR method is useful for estimating accuracies and bias in complex models. When the population size is small, old data are useful, and even a small amount of genomic information can substantially improve the accuracy. The effect of heat stress on first parity milk yield is small.

Evaluation of Genome-Enabled Prediction for Carcass Primal Cut Yields Using Single-Step Genomic Best Linear Unbiased Prediction in Hanwoo Cattle

There is a growing interest worldwide in genetically selecting high-value cut carcass weights, which allows for increased profitability in the beef cattle industry. Primal cut yields have been proposed as a potential indicator of cutability and overall carcass merit, and it is worthwhile to assess the prediction accuracies of genomic selection for these traits. This study was performed to compare the prediction accuracy obtained from a conventional pedigree-based BLUP (PBLUP) and a single-step genomic BLUP (ssGBLUP) method for 10 primal cut traits-bottom round, brisket, chuck, flank, rib, shank, sirloin, striploin, tenderloin, and top round-in Hanwoo cattle with the estimators of the linear regression method. The dataset comprised 3467 phenotypic observations for the studied traits and 3745 genotyped individuals with 43,987 single-nucleotide polymorphisms. In the partial dataset, the accuracies ranged from 0.22 to 0.30 and from 0.37 to 0.54 as evaluated using the PBLUP and ssGBLUP models, respectively. The accuracies of PBLUP and ssGBLUP with the whole dataset varied from 0.45 to 0.75 (average 0.62) and from 0.52 to 0.83 (average 0.71), respectively. The results demonstrate that ssGBLUP performed better than PBLUP averaged over the 10 traits, in terms of prediction accuracy, regardless of considering a partial or whole dataset. Moreover, ssGBLUP generally showed less biased prediction and a value of dispersion closer to 1 than PBLUP across the studied traits. Thus, the ssGBLUP seems to be more suitable for improving the accuracy of predictions for primal cut yields, which can be considered a starting point in future genomic evaluation for these traits in Hanwoo breeding practice.

Investigating the persistence of accuracy of genomic predictions over time in broilers

Accuracy of genomic predictions is an important component of the selection response. The objectives of this research were: 1) to investigate trends for prediction accuracies over time in a broiler population of accumulated phenotypes, genotypes, and pedigrees and 2) to test if data from distant generations are useful to maintain prediction accuracies in selection candidates. The data contained 820K phenotypes for a growth trait (GT), 200K for two feed efficiency traits (FE1 and FE2), and 42K for a carcass yield trait (CY). The pedigree included 1,252,619 birds hatched over 7 years, of which 154,318 from the last 4 years were genotyped. Training populations were constructed adding 1 year of data sequentially, persistency of accuracy over time was evaluated using predictions from birds hatched in the three generations following or in the years after the training populations. In the first generation, before genotypes became available for the training populations (first 3 years of data), accuracies remained almost stable with successive additions of phenotypes and pedigree to the accumulated dataset. The inclusion of 1 year of genotypes in addition to 4 years of phenotypes and pedigree in the training population led to increases in accuracy of 54% for GT, 76% for FE1, 110% for CY, and 38% for FE2; on average, 74% of the increase was due to genomics. Prediction accuracies declined faster without than with genomic information in the training populations. When genotypes were unavailable, the average decline in prediction accuracy across traits was 41% from the first to the second generation of validation, and 51% from the second to the third generation of validation. When genotypes were available, the average decline across traits was 14% from the first to the second generation of validation, and 3% from the second to the third generation of validation. Prediction accuracies in the last three generations were the same when the training population included 5 or 2 years of data, and a decrease of ~7% was observed when the training population included only 1 year of data. Training sets including genomic information provided an increase in accuracy and persistence of genomic predictions compared with training sets without genomic data. The two most recent years of pedigree, phenotypic, and genomic data were sufficient to maintain prediction accuracies in selection candidates. Similar conclusions were obtained using validation populations per year.

Impact of including the cause of missing records on genetic evaluations for growth in commercial pigs

It is of interest to evaluate crossbred pigs for hot carcass weight (HCW) and birth weight (BW); however, obtaining a HCW record is dependent on livability (LIV) and retained tag (RT). The purpose of this study is to analyze how HCW evaluations are affected when herd removal and missing identification are included in the model and examine if accounting for the reasons for missing traits improves the accuracy of predicting breeding values. Pedigree information was available for 1,965,077 purebred and crossbred animals. Records for 503,716 commercial three-way crossbred terminal animals from 2014 to 2019 were provided by Smithfield Premium Genetics. Two pedigree-based models were compared; model 1 (M1) was a threshold-linear model with all four traits (BW, HCW, RT, and LIV), and model 2 (M2) was a linear model including only BW and HCW. The fixed effects used in the model were contemporary group, sex, age at harvest (for HCW only), and dam parity. The random effects included direct additive genetic and random litter effects. Accuracy, dispersion, bias, and Pearson correlations were estimated using the linear regression method. The heritabilities were 0.11, 0.07, 0.02, and 0.04 for BW, HCW, RT, and LIV, respectively, with standard errors less than 0.01. No difference was observed in heritabilities or accuracies for BW and HCW between M1 and M2. Accuracies were 0.33, 0.37, 0.19, and 0.23 for BW, HCW, RT, and LIV, respectively. The genetic correlation between BW and RT was 0.34 ± 0.03, and between BW and LIV was 0.56 ± 0.03. Similarly, the genetic correlation between HCW and RT was 0.26 ± 0.04, and between HCW and LIV was 0.09 ± 0.05, respectively. The positive and moderate genetic correlations between BW and other traits imply a heavier BW resulted in a higher probability of surviving to harvest. Genetic correlations between HCW and other traits were lower due to the large quantity of missing records. Despite the heritable and correlated aspects of RT and LIV, results imply no major differences between M1 and M2; hence, it is unnecessary to include these traits in classical models for BW and HCW.

Multi-Trait Single-Step GBLUP Improves Accuracy of Genomic Prediction for Carcass Traits Using Yearling Weight and Ultrasound Traits in Hanwoo

There has been a growing interest in the genetic improvement of carcass traits as an important and primary breeding goal in the beef cattle industry over the last few decades. The use of correlated traits and molecular information can aid in obtaining more accurate estimates of breeding values. This study aimed to assess the improvement in the accuracy of genetic predictions for carcass traits by using ultrasound measurements and yearling weight along with genomic information in Hanwoo beef cattle by comparing four evaluation models using the estimators of the recently developed linear regression method. We compared the performance of single-trait pedigree best linear unbiased prediction [ST-BLUP and single-step genomic (ST-ssGBLUP)], as well as multi-trait (MT-BLUP and MT-ssGBLUP) models for the studied traits at birth and yearling date of steers. The data comprised of 15,796 phenotypic records for yearling weight and ultrasound traits as well as 5,622 records for carcass traits (backfat thickness, carcass weight, eye muscle area, and marbling score), resulting in 43,949 single-nucleotide polymorphisms from 4,284 steers and 2,332 bulls. Our results indicated that averaged across all traits, the accuracy of ssGBLUP models (0.52) was higher than that of pedigree-based BLUP (0.34), regardless of the use of single- or multi-trait models. On average, the accuracy of prediction can be further improved by implementing yearling weight and ultrasound data in the MT-ssGBLUP model (0.56) for the corresponding carcass traits compared to the ST-ssGBLUP model (0.49). Moreover, this study has shown the impact of genomic information and correlated traits on predictions at the yearling date (0.61) using MT-ssGBLUP models, which was advantageous compared to predictions at birth date (0.51) in terms of accuracy. Thus, using genomic information and high genetically correlated traits in the multi-trait model is a promising approach for practical genomic selection in Hanwoo cattle, especially for traits that are difficult to measure.

Genomic Prediction in Local Breeds: The Rendena Cattle as a Case Study

Simple Summary

Although genomic selection is being used in many livestock species, it has not yet been considered in local breeds due to the lower population size and the potential less effective impact on the genetic evaluation of these breeds. The current research aims to investigate how genomic data can impact the accuracy of genetic predictions for beef traits in Rendena, a small local cattle breed of the North-East of Italy selected for a dual purpose. Classical animal models using only phenotypic information were compared with two models that integrated genomic data with pedigree information. The genomic models presented better accuracy in estimated breeding values of the animals than the ‘classical’ animal model, especially the ‘simpler’ one assuming homogeneous variances of single nucleotide polymorphisms. Our results show that the inclusion of genomic information can be successfully applied to breeding selection scenarios even in small local cattle breeds such as Rendena.

Abstract

The maintenance of local cattle breeds is key to selecting for efficient food production, landscape protection, and conservation of biodiversity and local cultural heritage. Rendena is an indigenous cattle breed from the alpine North-East of Italy, selected for dual purpose, but with lesser emphasis given to beef traits. In this situation, increasing accuracy for beef traits could prevent detrimental effects due to the antagonism with milk production. Our study assessed the impact of genomic information on estimated breeding values (EBVs) in Rendena performance-tested bulls. Traits considered were average daily gain, in vivo EUROP score, and in vivo estimate of dressing percentage. The final dataset contained 1691 individuals with phenotypes and 8372 animals in pedigree, 1743 of which were genotyped. Using the cross-validation method, three models were compared: (i) Pedigree-BLUP (PBLUP); (ii) single-step GBLUP (ssGBLUP), and (iii) weighted single-step GBLUP (WssGBLUP). Models including genomic information presented higher accuracy, especially WssGBLUP. However, the model with the best overall properties was the ssGBLUP, showing higher accuracy than PBLUP and optimal values of bias and dispersion parameters. Our study demonstrated that integrating phenotypes for beef traits with genomic data can be helpful to estimate EBVs, even in a small local breed.

Genomic information allows for more accurate breeding values for milkability in dual-purpose Italian Simmental cattle

Milkability is a trait related to the milking efficiency of an animal, and it is a component of the herd profitability. Due to its economic importance, milkability is currently included in the selection index of the Italian Simmental cattle breed with a weight of 7.5%. This lowly heritable trait is measured on a subjective scale from 1 to 3 (1 = slow, 3 = fast), and genetic evaluations are performed by pedigree-based BLUP. Genomic information is now available for some animals in the Italian Simmental population, and its inclusion in the genetic evaluation system could increase accuracy of breeding values and genetic progress for milkability. The aim of this study was to test the feasibility and advantages of having a genomic evaluation for this trait in the Italian Simmental population. Phenotypes were available for 131,308 cows. A total of 9,526 animals had genotypes for 42,152 loci; among the genotyped animals, 2,455 were cows with phenotypes, and the other were their relatives. The youngest cows with both phenotypes and genotypes (n = 900) were identified as selection candidates. Variance components and heritability were estimated using pedigree information, whereas genetic and genomic evaluations were carried out using BLUP and single-step genomic BLUP (ssGBLUP), respectively. In addition, a weighted ssGBLUP was assessed using genomic regions from a genome-wide association study. Evaluation models were validated using theoretical and realized accuracies. The estimated heritability for milkability was 0.12 ± 0.01. The mean theoretical accuracies for selection candidates were 0.43 ± 0.08 (BLUP) and 0.53 ± 0.06 (ssGBLUP). The mean realized accuracies based on linear regression statistics were 0.29 (BLUP) and 0.40 (ssGBLUP). No genomic regions were significantly associated with milkability, thus no improvements in accuracy were observed when using weighted ssGBLUP. Results indicated that genomic information could improve the accuracy of breeding values and increase genetic progress for milkability in Italian Simmental.

Validation of single-step GBLUP genomic predictions from threshold models using the linear regression method: An application in chicken mortality

The objective of this study was to determine whether the linear regression (LR) method could be used to validate genomic threshold models. Statistics for the LR method were computed from estimated breeding values (EBVs) using the whole and truncated data sets with variances from the reference and validation populations. The method was tested using simulated and real chicken data sets. The simulated data set included 10 generations of 4,500 birds each; genotypes were available for the last three generations. Each animal was assigned a continuous trait, which was converted to a binary score assuming an incidence of failure of 7%. The real data set included the survival status of 186,596 broilers (mortality rate equal to 7.2%) and genotypes of 18,047 birds. Both data sets were analysed using best linear unbiased predictor (BLUP) or single-step GBLUP (ssGBLUP). The whole data set included all phenotypes available, whereas in the partial data set, phenotypes of the most recent generation were removed. In the simulated data set, the accuracies based on the LR formulas were 0.45 for BLUP and 0.76 for ssGBLUP, whereas the correlations between true breeding values and EBVs (i.e. true accuracies) were 0.37 and 0.65, respectively. The gain in accuracy by adding genomic information was overestimated by 0.09 when using the LR method compared to the true increase in accuracy. However, when the estimated ratio between the additive variance computed based on pedigree only and on pedigree and genomic information was considered, the difference between true and estimated gain was <0.02. Accuracies of BLUP and ssGBLUP with the real data set were 0.41 and 0.47, respectively. This small improvement in accuracy when using ssGBLUP with the real data set was due to population structure and lower heritability. The LR method is a useful tool for estimating improvements in accuracy of EBVs due to the inclusion of genomic information when traditional validation methods as k-fold validation and predictive ability are not applicable.