Direct Genomic Values Research Articles

Accuracies of genomic prediction of feed efficiency traits using different prediction and validation methods in an experimental Nelore cattle population.

Animal feeding is the most important economic component of beef production systems. Selection for feed efficiency has not been effective mainly due to difficult and high costs to obtain the phenotypes. The application of genomic selection using SNP can decrease the cost of animal evaluation as well as the generation interval. The objective of this study was to compare methods for genomic evaluation of feed efficiency traits using different cross-validation layouts in an experimental beef cattle population genotyped for a high-density SNP panel (BovineHD BeadChip assay 700k, Illumina Inc., San Diego, CA). After quality control, a total of 437,197 SNP genotypes were available for 761 Nelore animals from the Institute of Animal Science, Sertãozinho, São Paulo, Brazil. The studied traits were residual feed intake, feed conversion ratio, ADG, and DMI. Methods of analysis were traditional BLUP, single-step genomic BLUP (ssGBLUP), genomic BLUP (GBLUP), and a Bayesian regression method (BayesCπ). Direct genomic values (DGV) from the last 2 methods were compared directly or in an index that combines DGV with parent average. Three cross-validation approaches were used to validate the models: 1) YOUNG, in which the partition into training and testing sets was based on year of birth and testing animals were born after 2010; 2) UNREL, in which the data set was split into 3 less related subsets and the validation was done in each subset a time; and 3) RANDOM, in which the data set was randomly divided into 4 subsets (considering the contemporary groups) and the validation was done in each subset at a time. On average, the RANDOM design provided the most accurate predictions. Average accuracies ranged from 0.10 to 0.58 using BLUP, from 0.09 to 0.48 using GBLUP, from 0.06 to 0.49 using BayesCπ, and from 0.22 to 0.49 using ssGBLUP. The most accurate and consistent predictions were obtained using ssGBLUP for all analyzed traits. The ssGBLUP seems to be more suitable to obtain genomic predictions for feed efficiency traits on an experimental population of genotyped animals.

Phenotypic and genetic effects of recessive haplotypes on yield, longevity, and fertility

Phenotypes from the August 2015 US national genetic evaluation were used to compute phenotypic effects of 18 recessive haplotypes in Ayrshire (n=1), Brown Swiss (n=5), Holstein (n=10), and Jersey (n=2) cattle on milk, fat, and protein yields, somatic cell score (SCS), single-trait productive life (PL), daughter pregnancy rate (DPR), heifer conception rate (HCR), and cow conception rate (CCR). The haplotypes evaluated were Ayrshire haplotype 1, Brown Swiss haplotypes 1 and 2, spinal dysmyelination, spinal muscular atrophy, Weaver Syndrome, brachyspina, Holstein cholesterol deficiency, Holstein haplotypes 1 to 5, bovine leukocyte adhesion deficiency, complex vertebral malformation, mulefoot (syndactyly), and Jersey haplotypes 1 and 2. When causal variants are unknown and tests are based only on single nucleotide polymorphism haplotypes, it can sometimes be difficult to accurately determine carrier status. For example, 2 Holstein haplotypes for cholesterol deficiency have the same single nucleotide polymorphism genotype, but only one of them carries the causative mutation. Genotyped daughters of carrier bulls included in the analysis ranged from 8 for Weaver Syndrome to 17,869 for Holstein haplotype 3. Lactation records preadjusted for nongenetic factors and direct genomic values (DGV) were used to estimate phenotypic and genetic effects of recessive haplotypes, respectively. We found no phenotypic or genetic differences between carriers and noncarriers of Ayrshire or Brown Swiss defects. Several associations were noted for Holstein haplotypes, including fat and HCR for Holstein haplotype 0 carriers; milk, protein, SCS, PL, and fertility for Holstein haplotype 1; protein, PL, CCR, and HCR for Holstein haplotype 2; milk, protein, and fertility for Holstein haplotype 4; and protein yield and DPR for Holstein haplotype 5. There were no differences among bovine leukocyte adhesion deficiency carriers, but complex vertebral malformation affected fat yield and mulefoot carriers had higher SCS and lower PL DGV. Jersey haplotype 1 was not associated with any phenotypic effects, but we noted significant differences among DGV for fat, protein, PL, DPR, CCR, and HCR. Jersey haplotype 2 was associated only with lower phenotypic CCR. Effects of the recessive haplotypes on other traits studied generally were small even when significant. Almost $11 million of economic losses per year due to reduced fertility and perinatal calf death in the US population can be avoided by selecting mate pairs that will not produce affected embryos. Carrier animals may continue to be selected if the merit of their favorable alleles exceeds the loss from their recessive alleles, but carrier bulls can be generally avoided without reducing the average genetic merit of the sires available for mating.

Genomic clustering helps to improve prediction in a multibreed population.

Genomic prediction for crossbred beef cattle has shown limited results using low- to moderate-density SNP panels. The relationship between the training and validation populations, as well as the size of the reference population, affects the prediction accuracy for genomic selection. Rotational crossbreeding systems require the usage of crossbred animals as sires and dams of future generations, so crossbred animals require accurate evaluation. Here, a novel method for grouping of purebred and crossbred animals (based exclusively on genotypes) for genomic selection was investigated. Clustering of animals to investigate the genetic similarity among different groups was performed using several genomic relationship criteria between individuals. Hierarchical clusters based on average-link criteria (computed as the mean distance between elements of each subcluster) were formed. The accuracy of genomic prediction was assessed using 1,500 bulls genotyped for 54,609 markers. Estimated breeding values based on all available phenotypic records for birth weight, weaning gain, postweaning gain, and yearling gain were calculated using BLUP methodologies and deregressed to ensure unbiased comparisons could be made across populations. A 5-fold validation technique was used to calculate direct genomic values for all genotyped bulls; the addition of unrelated animals in the reference population was also investigated. We demonstrate a decrease in genomic selection accuracy after including animals from disconnected clusters. A method to improve genomic selection for crossbred and purebred animals by clustering animals based on their genotype is suggested. Unlike traditional approaches for genomic selection with a fixed reference population, genomic prediction using clusters (GPC) chooses the best reference population for better accuracy of genomic prediction of crossbred and purebred animals using clustering methods based on genotypes. An overall average gain in accuracy of 1.30% was noted over all scenarios across all traits investigated when the GPC approach was implemented. Further investigation is required to assess this difference in accuracy when a larger genotyped population is available, especially for the comparison of groups with higher genetic dissimilarity, such as those found in industry-wide across-breed genetic evaluations.

Accuracy of genomic prediction of purebreds for cross bred performance in pigs.

In pig breeding, as the final product is a cross bred (CB) animal, the goal is to increase the CB performance. This goal requires different strategies for the implementation of genomic selection from what is currently implemented in, for example dairy cattle breeding. A good strategy is to estimate marker effects on the basis of CB performance and subsequently use them to select pure bred (PB) breeding animals. The objective of our study was to assess empirically the predictive ability (accuracy) of direct genomic values of PB for CB performance across two traits using CB and PB genomic and phenotypic data. We studied three scenarios in which genetic merit was predicted within each population, and four scenarios where PB genetic merit for CB performance was predicted based on either CB or a PB training data. Accuracy of prediction of PB genetic merit for CB performance based on CB training data ranged from 0.23 to 0.27 for gestation length (GLE), whereas it ranged from 0.11 to 0.22 for total number of piglets born (TNB). When based on PB training data, it ranged from 0.35 to 0.55 for GLE and from 0.30 to 0.40 for TNB. Our results showed that it is possible to predict PB genetic merit for CB performance using CB training data, but predictive ability was lower than training using PB training data. This result is mainly due to the structure of our data, which had small-to-moderate size of the CB training data set, low relationship between the CB training and the PB validation populations, and a high genetic correlation (0.94 for GLE and 0.90 for TNB) between the studied traits in PB and CB individuals, thus favouring selection on the basis of PB data.

Technical note: Equivalent genomic models with a residual polygenic effect

Routine genomic evaluations in animal breeding are usually based on either a BLUP with genomic relationship matrix (GBLUP) or single nucleotide polymorphism (SNP) BLUP model. For a multi-step genomic evaluation, these 2 alternative genomic models were proven to give equivalent predictions for genomic reference animals. The model equivalence was verified also for young genotyped animals without phenotypes. Due to incomplete linkage disequilibrium of SNP markers to genes or causal mutations responsible for genetic inheritance of quantitative traits, SNP markers cannot explain all the genetic variance. A residual polygenic effect is normally fitted in the genomic model to account for the incomplete linkage disequilibrium. In this study, we start by showing the proof that the multi-step GBLUP and SNP BLUP models are equivalent for the reference animals, when they have a residual polygenic effect included. Second, the equivalence of both multi-step genomic models with a residual polygenic effect was also verified for young genotyped animals without phenotypes. Additionally, we derived formulas to convert genomic estimated breeding values of the GBLUP model to its components, direct genomic values and residual polygenic effect. Third, we made a proof that the equivalence of these 2 genomic models with a residual polygenic effect holds also for single-step genomic evaluation. Both the single-step GBLUP and SNP BLUP models lead to equal prediction for genotyped animals with phenotypes (e.g., reference animals), as well as for (young) genotyped animals without phenotypes. Finally, these 2 single-step genomic models with a residual polygenic effect were proven to be equivalent for estimation of SNP effects, too.

Can multi-subpopulation reference sets improve the genomic predictive ability for pigs?

In most countries and for most livestock species, genomic evaluations are obtained from within-breed analyses. To achieve reliable breeding values, however, a sufficient reference sample size is essential. To increase this size, the use of multibreed reference populations for small populations is considered a suitable option in other species. Over decades, the separate breeding work of different pig breeding organizations in Germany has led to stratified subpopulations in the breed German Large White. Due to this fact and the limited number of Large White animals available in each organization, there was a pressing need for ascertaining if multi-subpopulation genomic prediction is superior compared with within-subpopulation prediction in pigs. Direct genomic breeding values were estimated with genomic BLUP for the trait "number of piglets born alive" using genotype data (Illumina Porcine 60K SNP BeadChip) from 2,053 German Large White animals from five different commercial pig breeding companies. To assess the prediction accuracy of within- and multi-subpopulation reference sets, a random 5-fold cross-validation with 20 replications was performed. The five subpopulations considered were only slightly differentiated from each other. However, the prediction accuracy of the multi-subpopulations approach was not better than that of the within-subpopulation evaluation, for which the predictive ability was already high. Reference sets composed of closely related multi-subpopulation sets performed better than sets of distantly related subpopulations but not better than the within-subpopulation approach. Despite the low differentiation of the five subpopulations, the genetic connectedness between these different subpopulations seems to be too small to improve the prediction accuracy by applying multi-subpopulation reference sets. Consequently, resources should be used for enlarging the reference population within subpopulation, for example, by adding genotyped females.

Cross-validation of genetic and genomic predictions of temperament in Nellore–Angus crossbreds

The objective of this study was to cross-validate predictions of genomic merit for overall temperament at weaning (TEMP) by alternative partitioning of training and validation sets in a Nellore–Angus crossbred population. Nellore–Angus F2 calves (embryo transfer and natural service reciprocal crosses), paternal half-siblings to the embryo transfer calves, and F3 calves had records for TEMP (subjectively scored on a 1 to 9 scale). Calves with DNA available were genotyped on the Illumina Bovine SNP50 version 1 and 2 assays, and, after quality control filtering, there were 34,913 SNP markers available for use. Calves with TEMP records and genotypes (n=769; mean±std of all TEMP records is 3.975±2.062) were used in this study. BayesB procedures with π̃=0.997 were used to obtain direct genomic values (DGV) with alternative partitioning of data into training and validation sets utilizing the family structure of this population. Training was conducted (scenarios 1–4) using the progeny of all but one of the four sire or grandsires, (5) using only F2 progeny, (6) by random assignment, and (7) using only embryo transfer F2 and natural service half-sibling cattle. In random assignment, the number of animals included in the validation population was the average of scenarios 1–5. The DGV generated in these training and validation sets were compared to traditional, pedigree-based breeding values from an animal model with equivalent training and validation sets. The training model included fixed effects of sex, birth year-season combination, and temperament scoring pen nested in birth year-season combination. Standardized accuracies were higher for DGV using BayesB procedures (0.226 on average) compared to EBV using a traditional pedigree-based animal model (0.122 average). Random allocation of individuals into training and validation groups resulted in the highest accuracies for DGV (0.503) and EBV (0.354) of the validation groups. Overall, accuracies of animals in validation groups were low (i.e., less than 0.35) using genomic predictions, which could be due to inadequate sample size, insufficient marker density, or limited relationship between individuals in the training and validation populations. The genomic method generated unique DGV among full- and half-siblings within families whereas pedigree-based prediction results in identical EBV among full-siblings. This is of particular interest for the cattle industry as genomic methods can provide more plausible estimates of genetic merit for crossbred or purebred cattle without records, even if the initial accuracy is low.

Accuracy of genomic prediction using deregressed breeding values estimated from purebred and crossbred offspring phenotypes in pigs.

Genomic selection is applied to dairy cattle breeding to improve the genetic progress of purebred (PB) animals, whereas in pigs and poultry the target is a crossbred (CB) animal for which a different strategy appears to be needed. The source of information used to estimate the breeding values, i.e., using phenotypes of CB or PB animals, may affect the accuracy of prediction. The objective of our study was to assess the direct genomic value (DGV) accuracy of CB and PB pigs using different sources of phenotypic information. Data used were from 3 populations: 2,078 Dutch Landrace-based, 2,301 Large White-based, and 497 crossbreds from an F1 cross between the 2 lines. Two female reproduction traits were analyzed: gestation length (GLE) and total number of piglets born (TNB). Phenotypes used in the analyses originated from offspring of genotyped individuals. Phenotypes collected on CB and PB animals were analyzed as separate traits using a single-trait model. Breeding values were estimated separately for each trait in a pedigree BLUP analysis and subsequently deregressed. Deregressed EBV for each trait originating from different sources (CB or PB offspring) were used to study the accuracy of genomic prediction. Accuracy of prediction was computed as the correlation between DGV and the DEBV of the validation population. Accuracy of prediction within PB populations ranged from 0.43 to 0.62 across GLE and TNB. Accuracies to predict genetic merit of CB animals with one PB population in the training set ranged from 0.12 to 0.28, with the exception of using the CB offspring phenotype of the Dutch Landrace that resulted in an accuracy estimate around 0 for both traits. Accuracies to predict genetic merit of CB animals with both parental PB populations in the training set ranged from 0.17 to 0.30. We conclude that prediction within population and trait had good predictive ability regardless of the trait being the PB or CB performance, whereas using PB population(s) to predict genetic merit of CB animals had zero to moderate predictive ability. We observed that the DGV accuracy of CB animals when training on PB data was greater than or equal to training on CB data. However, when results are corrected for the different levels of reliabilities in the PB and CB training data, we showed that training on CB data does outperform PB data for the prediction of CB genetic merit, indicating that more CB animals should be phenotyped to increase the reliability and, consequently, accuracy of DGV for CB genetic merit.

A simple method to separate base population and segregation effects in genomic relationship matrices.

BackgroundGenomic selection and estimation of genomic breeding values (GBV) are widely used in cattle and plant breeding. Several studies have attempted to detect population subdivision by investigating the structure of the genomic relationship matrix G. However, the question of how these effects influence GBV estimation using genomic best linear unbiased prediction (GBLUP) has received little attention.MethodsWe propose a simple method to decompose G into two independent covariance matrices, one describing the covariance that results from systematic differences in allele frequencies between groups at the pedigree base (GA*) and the other describing genomic relationships (GS) corrected for these differences. Using this decomposition and Fst statistics, we examined whether observed genetic distances between genotyped subgroups within populations resulted from the heterogeneous genetic structure present at the base of the pedigree and/or from breed divergence. Using this decomposition, we tested three models in a forward prediction validation scenario on six traits using Brown Swiss and dual-purpose Fleckvieh cattle data. Model 0 (M0) used both components and is equivalent to the model using the standard G-matrix. Model 1 (M1) used GS only and model 2 (M2), an extension of M1, included a fixed genetic group effect. Moreover, we analyzed the matrix of contributions of each base group (Q) and estimated the effects and prediction errors of each base group using M0 and M1.ResultsThe proposed decomposition of G helped to examine the relative importance of the effects of base groups and segregation in a given population. We found significant differences between the effects of base groups for each breed. In forward prediction, differences between models in terms of validation reliability of estimated direct genomic values were small but predictive power was consistently lowest for M1. The relative advantage of M0 or M2 in prediction depended on breed, trait and genetic composition of the validation group. Our approach presents a general analogy with the use of genetic groups in conventional animal models and provides proof that standard GBLUP using G yields solutions equivalent to M0, where base groups are considered as correlated random effects within the additive genetic variance assigned to the genetic base.

Genome-wide association study for semen production traits in Holstein-Friesian bulls

Identifying genomic regions, particularly individual genes associated with semen quality traits, may be very important for improving sire fertility via selective breeding. The aim of the study was to estimate (co)variance components and effects of single nucleotide polymorphisms (SNP) from the Illumina BovineSNP50 BeadChip (Illumina, San Diego, CA) on semen production traits and to find candidate genes for these traits. The analyzed data set originates from the Polish Holstein-Friesian dairy cattle population and consists of 1,212 bulls kept at 4 artificial insemination stations. For each bull, 5 semen production traits were collected: sperm concentration, semen volume, number of spermatozoa, motility, and motility score. A multitrait mixed model was used to estimate genetic parameters. The parameters obtained were used to estimate SNP effects for each trait separately by the mixed model, which is used in the Polish direct genomic value project. Additionally, genes located in the vicinity of significant SNP were selected as candidate genes. For motility, 20 genome-wide significant SNP, located on 12 autosomes, were identified. For sperm concentration, we found 7 significant SNP: 3 on chromosome X, and 1 on chromosomes 1, 6, 23, and 24. For semen volume and motility score, 3 and 1 significant SNP were detected, respectively. All these SNP were located on chromosome X. For the number of spermatozoa, 12 significant SNP were observed. Six SNP were located on chromosome X, 3 on chromosome 8, and 1 on chromosomes 2, 7, and 16. This study clearly indicated a key role of the X chromosome in the determination of semen quality and emphasized that including such traits into genetic evaluation should be strongly considered.

Genetic evaluation using single-step genomic best linear unbiased predictor in American Angus.

Predictive ability of genomic EBV when using single-step genomic BLUP (ssGBLUP) in Angus cattle was investigated. Over 6 million records were available on birth weight (BiW) and weaning weight (WW), almost 3.4 million on postweaning gain (PWG), and over 1.3 million on calving ease (CE). Genomic information was available on, at most, 51,883 animals, which included high and low EBV accuracy animals. Traditional EBV was computed by BLUP and genomic EBV by ssGBLUP and indirect prediction based on SNP effects was derived from ssGBLUP; SNP effects were calculated based on the following reference populations: ref_2k (contains top bulls and top cows that had an EBV accuracy for BiW ≥0.85), ref_8k (contains all parents that were genotyped), and ref_33k (contains all genotyped animals born up to 2012). Indirect prediction was obtained as direct genomic value (DGV) or as an index of DGV and parent average (PA). Additionally, runs with ssGBLUP used the inverse of the genomic relationship matrix calculated by an algorithm for proven and young animals (APY) that uses recursions on a small subset of reference animals. An extra reference subset included 3,872 genotyped parents of genotyped animals (ref_4k). Cross-validation was used to assess predictive ability on a validation population of 18,721 animals born in 2013. Computations for growth traits used multiple-trait linear model and, for CE, a bivariate CE-BiW threshold-linear model. With BLUP, predictivities were 0.29, 0.34, 0.23, and 0.12 for BiW, WW, PWG, and CE, respectively. With ssGBLUP and ref_2k, predictivities were 0.34, 0.35, 0.27, and 0.13 for BiW, WW, PWG, and CE, respectively, and with ssGBLUP and ref_33k, predictivities were 0.39, 0.38, 0.29, and 0.13 for BiW, WW, PWG, and CE, respectively. Low predictivity for CE was due to low incidence rate of difficult calving. Indirect predictions with ref_33k were as accurate as with full ssGBLUP. Using the APY and recursions on ref_4k gave 88% gains of full ssGBLUP and using the APY and recursions on ref_8k gave 97% gains of full ssGBLUP. Genomic evaluation in beef cattle with ssGBLUP is feasible while keeping the models (maternal, multiple trait, and threshold) already used in regular BLUP. Gains in predictivity are dependent on the composition of the reference population. Indirect predictions via SNP effects derived from ssGBLUP allow for accurate genomic predictions on young animals, with no advantage of including PA in the index if the reference population is large. With the APY conditioning on about 10,000 reference animals, ssGBLUP is potentially applicable to a large number of genotyped animals without compromising predictive ability.

Genomic prediction for tick resistance in Braford and Hereford cattle.

One of the main animal health problems in tropical and subtropical cattle production is the bovine tick, which causes decreased performance, hide devaluation, increased production costs with acaricide treatments, and transmission of infectious diseases. This study investigated the utility of genomic prediction as a tool to select Braford (BO) and Hereford (HH) cattle resistant to ticks. The accuracy and bias of different methods for direct and blended genomic prediction was assessed using 10,673 tick counts obtained from 3,435 BO and 928 HH cattle belonging to the Delta G Connection breeding program. A subset of 2,803 BO and 652 HH samples were genotyped and 41,045 markers remained after quality control. Log transformed records were adjusted by a pedigree repeatability model to estimate variance components, genetic parameters, and breeding values (EBV) and subsequently used to obtain deregressed EBV. Estimated heritability and repeatability for tick counts were 0.19 ± 0.03 and 0.29 ± 0.01, respectively. Data were split into 5 subsets using k-means and random clustering for cross-validation of genomic predictions. Depending on the method, direct genomic value (DGV) prediction accuracies ranged from 0.35 with Bayes least absolute shrinkage and selection operator (LASSO) to 0.39 with BayesB for k-means clustering and between 0.42 with BayesLASSO and 0.45 with BayesC for random clustering. All genomic methods were superior to pedigree BLUP (PBLUP) accuracies of 0.26 for k-means and 0.29 for random groups, with highest accuracy gains obtained with BayesB (39%) for k-means and BayesC (55%) for random groups. Blending of historical phenotypic and pedigree information by different methods further increased DGV accuracies by values between 0.03 and 0.05 for direct prediction methods. However, highest accuracy was observed with single-step genomic BLUP with values of 0.48 for -means and 0.56, which represent, respectively, 84 and 93% improvement over PBLUP. Observed random clustering cross-validation breed-specific accuracies ranged between 0.29 and 0.36 for HH and between 0.55 and 0.61 for BO, depending on the blending method. These moderately high values for BO demonstrate that genomic predictions could be used as a practical tool to improve genetic resistance to ticks and in the development of resistant lines of this breed. For HH, accuracies are still in the low to moderate side and this breed training population needs to be increased before genomic selection could be reliably applied to improve tick resistance.

How imputation errors bias genomic predictions

The objective of this study was to investigate in detail the biasing effects of imputation errors on genomic predictions. Direct genomic values (DGV) of 3,494 Brown Swiss selection candidates for 37 production and conformation traits were predicted using either their observed 50K genotypes or their 50K genotypes imputed from a mimicked 6K chip. Changes in DGV caused by imputation errors were shown to be systematic. The DGV of top animals were, on average, underestimated and that of bottom animals were, on average, overestimated when imputed genotypes were used instead of observed genotypes. This pattern might be explained by the fact that imputation algorithms will usually suggest the most frequent haplotype from the sample whenever a haplotype cannot be determined unambiguously. That was empirically shown to cause an advantage for the bottom animals and a disadvantage for the top animals.

Including overseas performance information in genomic evaluations of Australian dairy cattle

In dairy cattle, the rate of genetic gain from genomic selection depends on reliability of direct genomic values (DGV). One option to increase reliabilities could be to increase the size of the reference set used for prediction, by using genotyped bulls with daughter information in countries other than the evaluating country. The increase in reliabilities of DGV from using this information will depend on the extent of genotype by environment interaction between the evaluating country and countries contributing information, and whether this is correctly accounted for in the prediction method. As the genotype by environment interaction between Australia and Europe or North America is greater than between Europe and North America for most dairy traits, ways of including information from other countries in Australian genomic evaluations were examined. Thus, alternative approaches for including information from other countries and their effect on the reliability and bias of DGV of selection candidates were assessed. We also investigated the effect of including overseas (OS) information on reliabilities of DGV for selection candidates that had weaker relationships to the current Australian reference set. The DGV were predicted either using daughter trait deviations (DTD) for the bulls with daughters in Australia, or using this information as well as OS information by including deregressed proofs (DRP) from Interbull for bulls with only OS daughters in either single trait or bivariate models. In the bivariate models, DTD and DRP were considered as different traits. Analyses were performed for Holstein and Jersey bulls for milk yield traits, fertility, cell count, survival, and some type traits. For Holsteins, the data used included up to 3,580 bulls with DTD and up to 5,720 bulls with only DRP. For Jersey, about 900 bulls with DTD and 1,820 bulls with DRP were used. Bulls born after 2003 and genotyped cows that were not dams of genotyped bulls were used for validation. The results showed that the combined use of DRP on bulls with OS daughters only and DTD for Australian bulls in either the single trait or bivariate model increased the coefficient of determination [(R2) (DGV,DTD)] in the validation set, averaged across 6 main traits, by 3% in Holstein and by 5% in Jersey validation bulls relative to the use of DTD only. Gains in reliability and unbiasedness of DGV were similar for the single trait and bivariate models for production traits, whereas the bivariate model performed slightly better for somatic cell count in Holstein. The increase in R2 (DGV,DTD) as a result of using bulls with OS daughters was relatively higher for those bulls and cows in the validation sets that were less related to the current reference set. For example, in Holstein, the average increase in R2 for milk yield traits when DTD and DRP were used in a single trait model was 23% in the least-related cow group, but only 3% in the most-related cow group. In general, for both breeds the use of DTD from domestic sources and DRP from Interbull in a single trait or bivariate model can increase reliability of DGV for selection candidates.

Multiple-breed genomic evaluation by principal component analysis in small size populations

In this study, the effects of breed composition and predictor dimensionality on the accuracy of direct genomic values (DGV) in a multiple breed (MB) cattle population were investigated. A total of 3559 bulls of three breeds were genotyped at 54 001 single nucleotide polymorphisms: 2093 Holstein (H), 749 Brown Swiss (B) and 717 Simmental (S). DGV were calculated using a principal component (PC) approach for either single (SB) or MB scenarios. Moreover, DGV were computed using all SNP genotypes simultaneously with SNPBLUP model as comparison. A total of seven data sets were used: three with a SB each, three with different pairs of breeds (HB, HS and BS), and one with all the three breeds together (HBS), respectively. Editing was performed separately for each scenario. Reference populations differed in breed composition, whereas the validation bulls were the same for all scenarios. The number of SNPs retained after data editing ranged from 36 521 to 41 360. PCs were extracted from actual genotypes. The total number of retained PCs ranged from 4029 to 7284 in Brown Swiss and HBS respectively, reducing the number of predictors by about 85% (from 82% to 89%). In all, three traits were considered: milk, fat and protein yield. Correlations between deregressed proofs and DGV were used to assess prediction accuracy in validation animals. In the SB scenarios, average DGV accuracy did not substantially change when either SNPBLUP or PC were used. Improvement of DGV accuracy were observed for some traits in Brown Swiss, only when MB reference populations and PC approach were used instead of SB-SNPBLUP (+10% HBS, +16%HB for milk yield and +3% HBS and +7% HB for protein yield, respectively). With the exclusion of the abovementioned cases, similar accuracies were observed using MB reference population, under the PC or SNPBLUP models. Random variation owing to sampling effect or size and composition of the reference population may explain the difficulty in finding a defined pattern in the results.

Comparison of breeding value prediction for two traits in a Nellore-Angus crossbred population using different Bayesian modeling methodologies.

The objectives of this study were to 1) compare four models for breeding value prediction using genomic or pedigree information and 2) evaluate the impact of fixed effects that account for family structure. Comparisons were made in a Nellore-Angus population comprising F2, F3 and half-siblings to embryo transfer F2 calves with records for overall temperament at weaning (TEMP; n = 769) and Warner-Bratzler shear force (WBSF; n = 387). After quality control, there were 34,913 whole genome SNP markers remaining. Bayesian methods employed were BayesB (π̃ = 0.995 or 0.997 for WBSF or TEMP, respectively) and BayesC (π = 0 and π̃), where π̃ is the ideal proportion of markers not included. Direct genomic values (DGV) from single trait Bayesian analyses were compared to conventional pedigree-based animal model breeding values. Numerically, BayesC procedures (using π̃) had the highest accuracy of all models for WBSF and TEMP (ρ̂gĝ = 0.843 and 0.923, respectively), but BayesB had the least bias (regression of performance on prediction closest to 1, β̂y,x = 2.886 and 1.755, respectively). Accounting for family structure decreased accuracy and increased bias in prediction of DGV indicating a detrimental impact when used in these prediction methods that simultaneously fit many markers.

XVI(th) QTLMAS: simulated dataset and comparative analysis of submitted results for QTL mapping and genomic evaluation.

BackgroundA common dataset was simulated and made available to participants of the XVIth QTL-MAS workshop. Tasks for the participants were to detect QTLs affecting three traits, to assess their possible pleiotropic effects, and to evaluate the breeding values in a candidate population without phenotypes using genomic information.MethodsFour generations consisting of 20 males and 1000 females were generated by mating each male with 50 females. The genome consisted of 5 chromosomes, each of 100 Mb size and carrying 2,000 equally distributed SNPs. Three traits were simulated in order to mimic milk yield, fat yield and fat content. Genetic (co)variances were generated from 50 QTLs with pleiotropic effects. Phenotypes for all traits were expressed only in females, and were provided for the first 3 generations. Fourteen methods for detecting single-trait QTL and 3 methods for investigating their pleiotropic nature were proposed. QTL mapping results were compared according to the following criteria: number of true QTL detected; number of false positives; and the proportion of the true genetic variance explained by submitted positions. Eleven methods for estimating direct genomic values of the candidate population were proposed. Accuracies and bias of predictions were assessed by comparing estimated direct genomic values with true breeding values.ResultsThe number of true detections ranged from 0 to 8 across methods and traits, false positives from 0 to 15, and the proportion of genetic variance captured from 0 to 0.82, respectively. The accuracy and bias of genomic predictions varied from 0.74 to 0.85 and from 0.86 to 1.34 across traits and methods, respectively.ConclusionsThe best results in terms of detection power were obtained by ridge regression that, however, led to a large number of false positives. Good results both in terms of true detections and false positives were obtained by the approaches that fit polygenic effects in the model. The investigation of the pleiotropic nature of the QTL permitted the identification of few additional markers compared to the single-trait analyses. Bayesian and grouped regularized regression methods performed similarly for genomic prediction while GBLUP produced the poorest results.

Breakeven prices for recording of indicator traits to reduce the environmental impact of milk production.

A breeding scheme using genomic selection and an indicator trait for environmental impact (EI) was studied to find the most effective recording strategy in terms of annual monetary genetic gain and breakeven price for the recording of indicator traits. The breakeven price shows the investment space for developing a recording system for an indicator trait. The breeding goal consisted of three traits – milk production, functional trait and environmental impact – with economic values of €83, €82 and €-83, respectively. The first scenario included only breeding goal traits and no indicator traits (NoIT). The other scenarios included all three breeding goal traits and one indicator trait (IT) for EI. The indicator traits were recorded on a large scale (stayability after first lactation and stature), medium scale (live weight and greenhouse gases (GHG) measured in the breath of the cow during milking) or small scale (residual feed intake and total enteric methane measured in a respiration chamber). In the scenario with stayability, the genetic gain in EI was over 11% higher than it was in NoIT. The breakeven price of recording stayability was €8 per record. Stayability is easy to record in the national milk recording system, and its use as an indicator trait for EI would not generate any additional recording costs. Therefore, stayability would be a good indicator trait to use to mitigate EI. The highest genetic gain in EI (23% higher compared to NoIT) was achieved when the GHG measured in the breath of the cow was used as indicator trait. The breakeven price for this indicator trait was €29 per record in the reference population. Ideally the recording of a specific indicator trait for EI would take place when: (i) the genetic correlation between the IT and EI is high; and (ii) the number of phenotypic records for the indicator trait is high enough to achieve a moderately high reliability of direct genomic values.

Impact of genotype imputation on the performance of GBLUP and Bayesian methods for genomic prediction.

The aim of this study was to evaluate the impact of genotype imputation on the performance of the GBLUP and Bayesian methods for genomic prediction. A total of 10,309 Holstein bulls were genotyped on the BovineSNP50 BeadChip (50 k). Five low density single nucleotide polymorphism (SNP) panels, containing 6,177, 2,480, 1,536, 768 and 384 SNPs, were simulated from the 50 k panel. A fraction of 0%, 33% and 66% of the animals were randomly selected from the training sets to have low density genotypes which were then imputed into 50 k genotypes. A GBLUP and a Bayesian method were used to predict direct genomic values (DGV) for validation animals using imputed or their actual 50 k genotypes. Traits studied included milk yield, fat percentage, protein percentage and somatic cell score (SCS). Results showed that performance of both GBLUP and Bayesian methods was influenced by imputation errors. For traits affected by a few large QTL, the Bayesian method resulted in greater reductions of accuracy due to imputation errors than GBLUP. Including SNPs with largest effects in the low density panel substantially improved the accuracy of genomic prediction for the Bayesian method. Including genotypes imputed from the 6 k panel achieved almost the same accuracy of genomic prediction as that of using the 50 k panel even when 66% of the training population was genotyped on the 6 k panel. These results justified the application of the 6 k panel for genomic prediction. Imputations from lower density panels were more prone to errors and resulted in lower accuracy of genomic prediction. But for animals that have close relationship to the reference set, genotype imputation may still achieve a relatively high accuracy.

Methods for genomic evaluation of a relatively small genotyped dairy population and effect of genotyped cow information in multiparity analyses

Methods for genomic prediction were evaluated for an Israeli Holstein dairy population of 713,686 cows and 1,305 progeny-tested bulls with genotypes. Inclusion of genotypes of 343 elite cows in an evaluation method that considers pedigree, phenotypes, and genotypes simultaneously was also evaluated. Two data sets were available: a complete data set with production records from 1985 through 2011, and a reduced data set with records after 2006 deleted. For each production trait, a multitrait animal model was used to compute traditional genetic evaluations for parities 1 through 3 as separate traits. Evaluations were calculated for the reduced and complete data sets. The evaluations from the reduced data set were used to calculate parent average for validation bulls, which was the benchmark for comparing gain in predictive ability from genomics. Genomic predictions for bulls in 2006 were calculated using a Bayesian regression method (BayesC), genomic BLUP (GBLUP), single-step GBLUP (ssGBLUP), and weighted ssGBLUP (WssGBLUP). Predictions using BayesC and GBLUP were calculated either with or without an index that included parent average. Genomic predictions that included elite cow genotypes were calculated using ssGBLUP and WssGBLUP. Predictive ability was assessed by coefficients of determination (R2) and regressions of predictions of 135 validation bulls with no daughters in 2006 on deregressed evaluations of those bulls in 2011. A reduction in R2 and regression coefficients was observed from parities 1 through 3. Fat and protein yields had the lowest R2 for all the methods. On average, R2 was lowest for parent averages, followed by GBLUP, BayesC, ssGBLUP, and WssGBLUP. For some traits, R2 for direct genomic values from BayesC and GBLUP were lower than those for parent averages. Genomic estimated breeding values using ssGBLUP were the least biased, and this method appears to be a suitable tool for genomic evaluation of a small genotyped population, as it automatically accounts for parental index, allows for inclusion of female genomic information without preadjustments in evaluations, and uses the same model as in traditional evaluations. Weighted ssGBLUP has the potential for higher evaluation accuracy.