Vincent Ducrocq

A common reference population from four European Holstein populations increases reliability of genomic predictions.

BackgroundSize of the reference population and reliability of phenotypes are crucial factors influencing the reliability of genomic predictions. It is therefore useful to combine closely related populations. Increased accuracies of genomic predictions depend on the number of individuals added to the reference population, the reliability of their phenotypes, and the relatedness of the populations that are combined.MethodsThis paper assesses the increase in reliability achieved when combining four Holstein reference populations of 4000 bulls each, from European breeding organizations, i.e. UNCEIA (France), VikingGenetics (Denmark, Sweden, Finland), DHV-VIT (Germany) and CRV (The Netherlands, Flanders). Each partner validated its own bulls using their national reference data and the combined data, respectively.ResultsCombining the data significantly increased the reliability of genomic predictions for bulls in all four populations. Reliabilities increased by 10%, compared to reliabilities obtained with national reference populations alone, when they were averaged over countries and the traits evaluated. For different traits and countries, the increase in reliability ranged from 2% to 19%.ConclusionsGenomic selection programs benefit greatly from combining data from several closely related populations into a single large reference population.

Rainbow trout resistance to bacterial cold-water disease is moderately heritable and is not adversely correlated with growth.

The objectives of this study were to estimate the heritabilities for and genetic correlations among resistance to bacterial cold-water disease and growth traits in a population of rainbow trout (Oncorhynchus mykiss). Bacterial cold-water disease, a chronic disease of rainbow trout, is caused by Flavobacterium psychrophilum. This bacterium also causes acute losses in young fish, known as rainbow trout fry syndrome. Selective breeding for increased disease resistance is a promising strategy that has not been widely used in aquaculture. At the same time, improving growth performance is critical for efficient production. At the National Center for Cool and Cold Water Aquaculture, reducing the negative impact of diseases on rainbow trout culture and improving growth performance are primary objectives. In 2005, when fish averaged 2.4 g, 71 full-sib families were challenged with F. psychrophilum and evaluated for 21 d. Overall survival was 29.3% and family rates of survival varied from 1.5 to 72.5%. Heritability of postchallenge survival, an indicator of disease resistance, was estimated to be 0.35 +/- 0.09. Body weights at 9 and 12 mo posthatch and growth rate from 9 to 12 mo were evaluated on siblings of the fish in the disease challenge study. Growth traits were moderately heritable, from 0.32 for growth rate to 0.61 for 12-mo BW. Genetic and phenotypic correlations between growth traits and resistance to bacterial cold-water disease were not different from zero. These results suggest that genetic improvement can be made simultaneously for growth and bacterial cold-water disease resistance in rainbow trout by using selective breeding.

Relationships between type and longevity in the Holstein breed

The relationship between type traits and longevity was studied in the French Holstein breed using a survival analysis model. In this model, the phenotypic value adjusted for systematic fixed effects, the estimated breeding value, or the residual value (defined as the difference between the adjusted phenotypic value and the estimated breeding value) of the cow for each type trait was included as a risk factor. This was done separately for two subpopulations (registered and nonregistered herds) and with or without adjustment for production traits, i.e., considering true or functional longevity. For both types of herds, udder traits (and above all, udder depth) clearly influenced the length of productive life. There seemed to be a more pronounced voluntary culling on type traits in registered herds. The correction for the within herd-year class of production traits, as a way to approximate functional longevity, increased the importance of udder traits and decreased the weight of capacity traits. The same results were obtained when the phenotypic value of the cow for type was replaced by her estimated breeding value, whereas residuals had little impact. The relationship between longevity and type traits was most often nonlinear, in particular for udder traits, but in this study, no trait with a clear intermediate optimum was found.

Effect of imputing markers from a low-density chip on the reliability of genomic breeding values in Holstein populations.

The purpose of this study was to investigate the imputation error and loss of reliability of direct genomic values (DGV) or genomically enhanced breeding values (GEBV) when using genotypes imputed from a 3,000-marker single nucleotide polymorphism (SNP) panel to a 50,000-marker SNP panel. Data consisted of genotypes of 15,966 European Holstein bulls from the combined EuroGenomics reference population. Genotypes with the low-density chip were created by erasing markers from 50,000-marker data. The studies were performed in the Nordic countries (Denmark, Finland, and Sweden) using a BLUP model for prediction of DGV and in France using a genomic marker-assisted selection approach for prediction of GEBV. Imputation in both studies was done using a combination of the DAGPHASE 1.1 and Beagle 2.1.3 software. Traits considered were protein yield, fertility, somatic cell count, and udder depth. Imputation of missing markers and prediction of breeding values were performed using 2 different reference populations in each country: either a national reference population or a combined EuroGenomics reference population. Validation for accuracy of imputation and genomic prediction was done based on national test data. Mean imputation error rates when using national reference animals was 5.5 and 3.9% in the Nordic countries and France, respectively, whereas imputation based on the EuroGenomics reference data set gave mean error rates of 4.0 and 2.1%, respectively. Prediction of GEBV based on genotypes imputed with a national reference data set gave an absolute loss of 0.05 in mean reliability of GEBV in the French study, whereas a loss of 0.03 was obtained for reliability of DGV in the Nordic study. When genotypes were imputed using the EuroGenomics reference, a loss of 0.02 in mean reliability of GEBV was detected in the French study, and a loss of 0.06 was observed for the mean reliability of DGV in the Nordic study. Consequently, the reliability of DGV using the imputed SNP data was 0.38 based on national reference data, and 0.48 based on EuroGenomics reference data in the Nordic validation, and the reliability of GEBV using the imputed SNP data was 0.41 based on national reference data, and 0.44 based on EuroGenomics reference data in the French validation.

Genomic selection in French dairy cattle

Genomic selection is implemented in French Holstein, Montbeliarde, and Normande breeds (70%, 16% and 12% of French dairy cows). A characteristic of the model for genomic evaluation is the use of haplotypes instead of single-nucleotide polymorphisms (SNPs), so as to maximise linkage disequilibrium between markers and quantitative

High-density marker imputation accuracy in sixteen French cattle breeds

BackgroundGenotyping with the medium-density Bovine SNP50 BeadChip® (50K) is now standard in cattle. The high-density BovineHD BeadChip®, which contains 777 609 single nucleotide polymorphisms (SNPs), was developed in 2010. Increasing marker density increases the level of linkage disequilibrium between quantitative trait loci (QTL) and SNPs and the accuracy of QTL localization and genomic selection. However, re-genotyping all animals with the high-density chip is not economically feasible. An alternative strategy is to genotype part of the animals with the high-density chip and to impute high-density genotypes for animals already genotyped with the 50K chip. Thus, it is necessary to investigate the error rate when imputing from the 50K to the high-density chip.MethodsFive thousand one hundred and fifty three animals from 16 breeds (89 to 788 per breed) were genotyped with the high-density chip. Imputation error rates from the 50K to the high-density chip were computed for each breed with a validation set that included the 20% youngest animals. Marker genotypes were masked for animals in the validation population in order to mimic 50K genotypes. Imputation was carried out using the Beagle 3.3.0 software.ResultsMean allele imputation error rates ranged from 0.31% to 2.41% depending on the breed. In total, 1980 SNPs had high imputation error rates in several breeds, which is probably due to genome assembly errors, and we recommend to discard these in future studies. Differences in imputation accuracy between breeds were related to the high-density-genotyped sample size and to the genetic relationship between reference and validation populations, whereas differences in effective population size and level of linkage disequilibrium showed limited effects. Accordingly, imputation accuracy was higher in breeds with large populations and in dairy breeds than in beef breeds. More than 99% of the alleles were correctly imputed if more than 300 animals were genotyped at high-density. No improvement was observed when multi-breed imputation was performed.ConclusionIn all breeds, imputation accuracy was higher than 97%, which indicates that imputation to the high-density chip was accurate. Imputation accuracy depends mainly on the size of the reference population and the relationship between reference and target populations.

An improved model for the French genetic evaluation of dairy bulls on length of productive life of their daughters

Functional longevity of dairy cows has been routinely evaluated in France since 1997 using a survival analysis model. Recently, we proposed a genetic trend validation test that could be used before including national data in an international evaluation of bulls on longevity of their daughters. Its application to the French Holstein data revealed a large overestimation of the genetic trend. It was found that the bias is the result of a change in the baseline hazard rate over time. A new proportional hazards model is proposed which accounts for this change. In the new model, the baseline is described as a stratified, piecewise Weibull hazard function within lactation, i.e. a function of the number of days since the most recent calving. Stratification is within year and parity. Different Weibull hazard functions are used over four periods: 0 to 270 days, 271 to 380 days, 381 days to day when dried, dry period until the next calving. The non-genetic effects included in the model were slightly different from the previous one. In particular the interaction effects between the within herd-year class of production and lactation number × stage of lactation on the one hand and year-season were accounted for. The estimated genetic variance was smaller than with the old model. The new genetic trend is almost flat. An illustration of the efficiency of selection on the estimated breeding values for longevity is presented.

Evidence of biases in genetic evaluations due to genomic preselection in dairy cattle

A genomic preselection step of young sires is now often included in dairy cattle breeding schemes. Young sires are selected based on their genomic breeding values. They have better Mendelian sampling contribution so that the assumption of random Mendelian sampling term in genetic evaluations is clearly violated. When these sires and their progeny are evaluated using BLUP, it is feared that estimated breeding values are biased. The effect of genomic selection on genetic evaluations was studied through simulations keeping the structure of the Holstein population in France. The quality of genetic evaluations was assessed by computing bias and accuracy from the difference and correlation between true and estimated breeding values, respectively, and also the mean square error of prediction. Different levels of heritability, selection intensity, and accuracy of genomic evaluation were tested. After only one generation and whatever the scenario, breeding values of preselected young sires and their daughters were significantly underestimated and their accuracy was decreased. Genomic preselection needs to be accounted for in genetic evaluation models.

Estimation of genetic parameters for test day records of dairy traits in the first three lactations

Application of test-day models for the genetic evaluation of dairy populations requires the solution of large mixed model equations. The size of the (co)variance matrices required with such models can be reduced through the use of its first eigenvectors. Here, the first two eigenvectors of (co)variance matrices estimated for dairy traits in first lactation were used as covariables to jointly estimate genetic parameters of the first three lactations. These eigenvectors appear to be similar across traits and have a biological interpretation, one being related to the level of production and the other to persistency. Furthermore, they explain more than 95% of the total genetic variation. Variances and heritabilities obtained with this model were consistent with previous studies. High correlations were found among production levels in different lactations. Persistency measures were less correlated. Genetic correlations between second and third lactations were close to one, indicating that these can be considered as the same trait. Genetic correlations within lactation were high except between extreme parts of the lactation. This study shows that the use of eigenvectors can reduce the rank of (co)variance matrices for the test-day model and can provide consistent genetic parameters.

Computational strategies for national integration of phenotypic, genomic, and pedigree data in a single-step best linear unbiased prediction

The single-step genomic BLUP (SSGBLUP) is a method that can integrate pedigree and genotypes at molecular markers in an optimal way. However, its present form (regular SSGBLUP) has a high computational cost (cubic in the number of genotyped animals) and may need extensive rewriting of genetic evaluation software. In this work, we propose several strategies to implement the single step in a simpler manner. The first one expands the single-step mixed-model equations to obtain equivalent equations from which the regular (including pedigree and records only) mixed-model equations are a subset. These new equations (unsymmetric extended SSGBLUP) have low computational cost, but require a nonsymmetric solver such as the biconjugate gradient stabilized method or successive underrelaxation, which is a variant of successive overrelaxation, with a relaxation factor lower than 1. In addition, we show a new derivation of the single-step method, which includes, as an extra effect, deviations from strictly polygenic breeding values. As a result, the same set of equations as above is obtained. We show that, whereas the new derivation shows apparent problems of nonpositive definiteness for certain covariance matrices, a proper equivalent model including imaginary effects always exists, leading always to the regular SSGBLUP mixed model equations. The system of equations can be solved (iterative SSGBLUP) by iterating between a pedigree and records evaluation and a genomic evaluation (each one solved by any iterative or direct method), whereas global iteration can use a block version of successive underrelaxation, which ensures convergence. The genomic evaluation can explicitly include marker or haplotype effects and possibly involve nonlinear (e.g., Bayesian by Markov chain Monte Carlo) methods. In a simulated example with 28,800 individuals and 1,800 genotyped individuals, all methods converged quickly to the same solutions. Using existing efficient methods with limited memory requirements to compute the products Gt and A(22)t for any t (where G and A(22) are genomic and pedigree relationships for genotyped animals, and t is a vector), all strategies can be converted to iteration on data procedures for which the total number of operations is linear in the number of animals + number of genotyped animals × number of markers.