Optimum contribution selection using traditional best linear unbiased prediction and genomic breeding values in aquaculture breeding schemes

The aim of this study was to compare genetic gain for a traditional aquaculture sib breeding scheme with breeding values based on phenotypic data (TBLUP) with a breeding scheme with genome-wide (GW) breeding values. Both breeding schemes were closed nuclei with discrete generations modeled by stochastic simulation. Optimum contribution selection was applied to restrict pedigree-based inbreeding to either 0.5 or 1% per generation. There were 1,000 selection candidates and a sib test group of either 4,000 or 8,000 fish. The number of selected dams and sires to create full sib families in each generation was determined from the optimum contribution selection method. True breeding values for a trait were simulated by summing the number of each QTL allele and the true effect of each of the 1,000 simulated QTL. Breeding values in TBLUP were predicted from phenotypic and pedigree information, whereas genomic breeding values were computed from genetic markers whose effects were estimated using a genomic BLUP model. In generation 5, genetic gain was 70 and 74% greater for the GW scheme than for the TBLUP scheme for inbreeding rates of 0.5 and 1%. The reduction in genetic variance was, however, greater for the GW scheme than for the TBLUP scheme due to fixation of some QTL. As expected, accuracy of selection increased with increasing heritability (e.g., from 0.77 with a heritability of 0.2 to 0.87 with a heritability of 0.6 for GW, and from 0.53 and 0.58 for TBLUP in generation 5 with sib information only). When the trait was measured on the selection candidate compared with only on sibs and the heritability was 0.4, accuracy increased from 0.55 to 0.69 for TBLUP and from 0.83 to 0.86 for GW. The number of selected sires to get the desired rate of inbreeding was in general less in GW than in TBLUP and was 33 for GW and 83 for TBLUP (rate of inbreeding 1% and heritability 0.4). With truncation selection, genetic gain for the scheme with GW breeding values was nearly twice as large as a scheme with traditional BLUP breeding values. The results indicate that the benefits of applying GW breeding values compared with TBLUP are reduced when contributions are optimized. In conclusion, genetic gain in aquaculture breeding schemes with optimized contributions can increase by as much as 81% by applying genome-wide breeding values compared with traditional BLUP breeding values.     

Two generations of selection on restricted best linear unbiased prediction breeding values for income minus feed cost in laying hens.

Two generations of selection on restricted BLUP breeding values were applied in an experiment with laying hens. Selection had been on phenotype of income minus feed cost (IFC) between 21 and 40 wk of age in the previous five generations. The restriction of no genetic change in egg weight was included in the EBV for power-transformed IFC (i.e., IFCt, with t-values of 3.7 and 3.6 in the two generations, respectively). The experiment consisted of two selection lines plus a randomly bred control of 20 male and 80 female breeders each. Observations on 8,844 survivors to 40 wk were available. Relative to the base population average, the restriction reduced genetic gain in IFC from 4.1 and 3.9% to 2.0 and 2.2% per generation in the two selection lines, respectively. Average EBV for egg weight remained nearly constant after a strong increase in the previous five generations. Rates of genetic gain for egg number, body weight, and feed conversion (feed/egg mass) were not affected significantly. In the seventh generation, a genetic gain in feed conversion of 10.3% relative to the phenotypic mean of the base population was obtained.     

Comparison of selection based on phenotype,selection index and best linear unbiased prediction using data from a closed broiler line

1. Selection based on three methods of estimating breeding values, Best Linear Unbiased Prediction (BLUP), selection index (SI), and phenotype (SP) were compared for three traits, juvenile body weight (JW), percentage breast meat yield (BM) and hen‐day rate of egg production (EP) using records provided by a commercial broiler breeding company.

2. Product moment correlations were calculated between breeding values estimated by each method and averaged across sexes. A mean correlation of 0.69 was obtained between selection on SP and BLUP for JW. Mean correlations of 0.88 and 0.68 and 0.87 were obtained between SI and BLUP for the traits JW, EP and BM, respectively.

3. A mean estimated genetic response of 77.7% was obtained with SP for JW relative to BLUP in the absence of restrictions on the selection of close relatives. Estimated genetic responses of 90.7%, 66.9% and 88.4% were obtained by SI relative to BLUP for JW, EP and BM, respectively.

4. Applying restrictions on the selection of close relatives resulted in slight decreases in estimated responses but not in the respective ranking of the selection methods.

5. The results indicate that BLUP could provide commercial breeders with increased selection responses compared to index selection, in particular for traits of low heritability and where relatively few animals possess performance records.     

A method of computing restricted best linear unbiased prediction of breeding values for some animals in a population

Restricted BLUP (R-BLUP) is derived by imposing restrictions directly within a multiple-trait mixed model. As a result, the R-BLUP procedure requires the solution of high-order simultaneous equations. If restrictions are imposed on breeding values for only some animals in a population, calculations become more complex. A new procedure for computing the R-BLUP of breeding values was derived when constraints were imposed on the additive genetic values of only some animals in a population. Rules for including records when proportional constraints are imposed were developed based on the traits that are recorded for an animal. The technique was better than the previous method in both memory requirement and central processing unit time.     

Efficacy of alternative multivariate best linear unbiased prediction models for genetic evaluation of swine.

A comparison of the accuracy of alternative BLUP evaluations of swine performance data is reported. A simulated data set of performance for days to market, backfat, number born alive, number weaned, and litter weight in six herds was used for the evaluation. The data structure was derived by using the performance records of six herds sampled from the American Yorkshire Club's STAGES (Swine Testing and Genetic Evaluation System) database. For growth traits, 10,360 pig records from 129 sires and 897 dams were recorded. For maternal traits, records on 2,598 litters of 1,209 sows from 147 sires and 585 dams were included. The actual observed performance of each record was removed and replaced with simulated performance. These simulated data were analyzed by within- and across-herd BLUP models using STAGES and Animal Model (AM) procedures. Results indicate that the alternative BLUP procedures produced similar estimates. Correlations between STAGES and AM EPD ranged from .84 to .95. Correlations between STAGES EPD and true genetic value (G) ranged from .41 to .74, and correlations between AM EPD and G ranged from .40 to .74. On average, AM EPD had a .04 larger correlation with G than did STAGES EPD, although the difference in the correlations was not significant (P greater than .05). Trends in EPD for sires, dams, and pigs or sows were the same. Likewise, standard errors of prediction for AM EPD averaged 4% smaller than those for STAGES EPD. Computationally, the AM procedures used 15 to 20 times as much processing time as did STAGES procedures.     

Detecting effective starting point of genomic selection by divergent trends from best linear unbiased prediction and single-step genomic best linear unbiased prediction in pigs,beef cattle,and broilers

Genomic selection has been adopted nationally and internationally in different livestock and plant species. However, understanding whether genomic selection has been effective or not is an essential question for both industry and academia. Once genomic evaluation started being used, estimation of breeding values with pedigree best linear unbiased prediction (BLUP) became biased because this method does not consider selection using genomic information. Hence, the effective starting point of genomic selection can be detected in two possible ways including the divergence of genetic trends and Realized Mendelian sampling (RMS) trends obtained with BLUP and single-step genomic BLUP (ssGBLUP). This study aimed to find the start date of genomic selection for a set of economically important traits in three livestock species by comparing trends obtained using BLUP and ssGBLUP. Three datasets were used for this purpose: 1) a pig dataset with 117k genotypes and 1.3M animals in pedigree, 2) an Angus cattle dataset consisted of ~842k genotypes and 11.5M animals in pedigree, and 3) a purebred broiler chicken dataset included ~154k genotypes and 1.3M birds in pedigree were used. The genetic trends for pigs diverged for the genotyped animals born in 2014 for average daily gain (ADG) and backfat (BF). In beef cattle, the trends started diverging in 2009 for weaning weight (WW) and in 2016 for postweaning gain (PWG), with little divergence for birth weight (BTW). In broiler chickens, the genetic trends estimated by ssGBLUP and BLUP diverged at breeding cycle 6 for two out of the three production traits. The RMS trends for the genotyped pigs diverged for animals born in 2014, more for ADG than for BF. In beef cattle, the RMS trends started diverging in 2009 for WW and in 2016 for PWG, with a trivial trend for BTW. In broiler chickens, the RMS trends from ssGBLUP and BLUP diverged strongly for two production traits at breeding cycle 6, with a slight divergence for another trait. Divergence of the genetic trends from ssGBLUP and BLUP indicates the onset of the genomic selection. The presence of trends for RMS indicates selective genotyping, with or without the genomic selection. The onset of genomic selection and genotyping strategies agrees with industry practices across the three species. In summary, the effective start of genomic selection can be detected by the divergence between genetic and RMS trends from BLUP and ssGBLUP.     

The effectiveness of the best linear unbiased prediction of beef sires using field data collected from small farms.

Beef sire evaluation using BLUP was investigated under field conditions at small-scale farms. The 6,848 records on fattened Japanese Black cattle steers were obtained from 1981 through 1987. The average number of steers in the subclass of market-year-farm was 5.2. A sire-maternal grandsire mixed model with relationships was used to analyze the data to yield BLUP for the sire and maternal grandsire effects. The regression coefficients of the realized value of progeny on the predicted value calculated using BLUP procedures for daily gain, carcass weight, and marbling score were 1.027, 1.054, and .917, respectively (i.e., the regression almost equaled 1). Therefore, BLUP was shown to be very effective in predicting offspring performance, even using field records collected from small-scale farms.     

An alternative derivation method of mixed model equations from best linear unbiased prediction (BLUP) and restricted BLUP of breeding values not using maximum likelihood

Two mixed model equations (MME) for best linear unbiased prediction (BLUP) of breeding values and for restricted BLUP of breeding values were derived by maximum likelihood from the joint normal probability distribution of the observations and breeding values. As a result, MME is actually more general than maximum likelihood because we can prove that each set of solutions of MME are identical to BLUP and restricted BLUP of breeding values and then it does not depend on normality. In the present study, the author shows deriving directly each MME from BLUP and restricted BLUP equations for breeding values without assuming the joint normal distribution of the data and random effects. However, if we cannot assume the multivariate normal density distribution of the estimated aggregate breeding value and each breeding value for selected traits, the response to selection by restricted BLUP may deviate from the expected values.     

Identifying single genes of large effect in quantitative traits using best linear unbiased prediction

The recent discovery of a major gene for rapid postweaning growth has reinforced the hypothesis that other quantitative continuous traits may be influenced by single genes of large effect. However, most methods for the detection of such genes rely on the discovery of multimodality in the population frequency distribution. The complicating effects of environment and artificial selection make the identification of such genes with field-collected data a formidable problem. An index is proposed that may serve as an indicator that a major gene is segregating within a population. The index is based on the assumption that under polygenic inheritance, an offspring's deviation from the midparent average is smaller than the deviation from either parent. Whereas, for the Mendelian segregation of a major gene, the opposite would be expected. A proposed class of indices is then based on the ratio [O - .5(S + D)]k/([O - S]k/2 [O - D]k/2) where O, S and D are the additive genetic values of an offspring and its sire and dam estimated via best linear unbiased prediction. Values of the index greater than 1. would be indicative of major gene inheritance. Simulation of small populations indicates that the index is quite sensitive to the existence of segregating major genes even in the absence of multimodality of the phenotypic distribution. However, the index remains dependent on the accuracy of genetic value estimation.     

Efficient selection against categorically scored hip dysplasia in dogs is possible using best linear unbiased prediction and optimum contribution selection: a simulation study

Breeding to reduce the prevalence of categorically scored hip dysplasia (HD), based on phenotypic assessment of radiographic hip status, has had limited success. The aim of this study was to evaluate two selection strategies for improved hip status: truncation selection based on phenotypic record versus best linear unbiased prediction (BLUP), using stochastic simulation and selection scenarios resembling those in real dog populations. In addition, optimum contribution selection (OCS) was evaluated. Two traits were considered: HD (as a categorical trait with five classes and a heritability of 0.45 on the liability scale) and a continuous trait (with a heritability of 0.25) intended to represent other characteristics in the breeding goal. A population structure mimicking that in real dog populations was modelled. The categorical nature of HD caused a considerably lower genetic gain compared to simulating HD as a continuous trait. Genetic gain was larger for BLUP selection than for phenotypic selection in all scenarios. However, BLUP selection resulted in higher rates of inbreeding. By applying OCS, the rate of inbreeding was lowered to about the same level as phenotypic selection but with increased genetic improvement. For efficient selection against HD, use of BLUP breeding values should be prioritized. In small populations, BLUP should be used together with OCS or similar strategy to maintain genetic variation.     

Effect of more stringent convergence criterion of estimated breeding values on response to selection

In the analysis of large amounts of data to obtain BLUP, large sets of mixed model equations must be solved iteratively, which can involve considerable computing time. In real life, solutions are required only periodically for breeders to choose the best individuals, so that computing time is not usually a limiting factor. In simulation studies involving evaluation of individuals by BLUP, many rounds of evaluation are required for each simulated population. Since several or many replicates are usually required to obtain an accurate result from stochastic simulations, computing time can become a limiting factor. One of the factors that can drastically affect computing time in iterative methods is the criterion for ceasing iteration, or convergence criterion (CC). With iterative methods, a disadvantage can be that the rate of convergence can be slow, or under certain circumstances not converge at all. Nevertheless, when the system converges, the more stringent the CC, the more accurate the solutions. The more stringent the CC the more iterations and hence more computing time is required. The objectives of this study were to investigate how much response to selection is affected by the stringency of the CC and how much reranking occurs among selected individuals at different levels of the convergence criteria. These explorations provided a profile analysis of the computing time spent for each of the major subroutines in the BBSIM program.     

Spread of pseudorabies virus among breeding swine in quarantined herds.

In theory, pseudorabies virus (PRV) may be eliminated from any size of breeding herd by phased test and removal if replacement gilts are not infected with PRV, culling decisions are partially based on PRV status, and the cull rate is higher than the incidence rate of PRV. Annual cull rates are commonly at least 50%, but little information exists on the incidence of PRV within enzootically infected swine herds. The purpose of this study was to develop a method by which spread of PRV could be detected among breeding swine within enzootically infected herds and to determine the incidence of PRV infection in these herds. Data were collected from 17 herds that were quarantined for PRV and ranged in size from 120 to 1,100 sows. At each herd, within the first 5 days of introduction, a group of approximately 30 replacement gilts was identified, vaccinated with a glycoprotein X-deleted PRV vaccine, and blood sample was collected. The owner of 1 herd had a nonvaccinated breeding herd and elected to leave incoming gilts nonvaccinated. After vaccination, blood samples were collected every 1 to 2 months for an average of 13.6 months. Serum samples from vaccinated gilts were tested for antiglycoprotein X antibodies by a specific differential ELISA. Samples from nonvaccinated gilts were evaluated by serum neutralization test. Product-limit method was used to estimate the probability of not becoming infected with PRV. Spread was detected in 7 of 8 herds that had more than 400 sows and in 2 of 9 herds that had less than 400 sows.(ABSTRACT TRUNCATED AT 250 WORDS)     

Evaluation of patterns of removal and associations among culling because of lameness and sow productivity traits in swine breeding herds

OBJECTIVE: To characterize patterns of removal and evaluate the associations among culling because of lameness and sow productivity traits among culled gilts and sows. DESIGN: Cross-sectional study. SAMPLE POPULATION: Data from a convenience sample of 11 farms pertaining to the removal of 51,795 gilts and sows from January 1991 to December 2002. Mean culling and mortality (death and euthanasia) rates for all inventoried gilts and sows ranged from 23% to 50% and 4.7% to 9.5%, respectively. PROCEDURE: An analysis of categories of removal (cull, death, or euthanasia) and reasons for removal of gilts and sows was performed. Multivariate logistic regression was used to determine associations among culling because of lameness and sow productivity traits among culled gilts and sows. RESULTS: Among sows that were removed, the proportion of parity > or = 1 sows that died (both death and euthanasia) was > 3 times the proportion of parity > or = 1 sows that were culled within 20 days after farrowing. Among lame sows that were removed, the proportion of parity > or = 1 sows that died (death and euthanasia) was higher than the proportion of parity > or = 1 sows that were culled within 20 days after farrowing. Among sows that were removed, the proportion of sows that died (deaths and euthanasia) was higher during lactation than nonlactation. This was also observed among lame sows that were removed. CONCLUSIONS AND CLINICAL RELEVANCE: The proportion of death among removed sows, especially lame sows, was higher during lactation than nonlactation. Results indicated that risk of death is not the same for sows throughout their lifetime.     

Core-dependent changes in genomic predictions using the Algorithm for Proven and Young in single-step genomic best linear unbiased prediction

Single-step genomic best linear unbiased prediction with the Algorithm for Proven and Young (APY) is a popular method for large-scale genomic evaluations. With the APY algorithm, animals are designated as core or noncore, and the computing resources to create the inverse of the genomic relationship matrix (GRM) are reduced by inverting only a portion of that matrix for core animals. However, using different core sets of the same size causes fluctuations in genomic estimated breeding values (GEBVs) up to one additive standard deviation without affecting prediction accuracy. About 2% of the variation in the GRM is noise. In the recursion formula for APY, the error term modeling the noise is different for every set of core animals, creating changes in breeding values. While average changes are small, and correlations between breeding values estimated with different core animals are close to 1.0, based on the normal distribution theory, outliers can be several times bigger than the average. Tests included commercial datasets from beef and dairy cattle and from pigs. Beyond a certain number of core animals, the prediction accuracy did not improve, but fluctuations decreased with more animals. Fluctuations were much smaller than the possible changes based on prediction error variance. GEBVs change over time even for animals with no new data as genomic relationships ties all the genotyped animals, causing reranking of top animals. In contrast, changes in nongenomic models without new data are small. Also, GEBV can change due to details in the model, such as redefinition of contemporary groups or unknown parent groups. In particular, increasing the fraction of blending of the GRM with a pedigree relationship matrix from 5% to 20% caused changes in GEBV up to 0.45 SD, with a correlation of GEBV > 0.99. Fluctuations in genomic predictions are part of genomic evaluation models and are also present without the APY algorithm when genomic evaluations are computed with updated data. The best approach to reduce the impact of fluctuations in genomic evaluations is to make selection decisions not on individual animals with limited individual accuracy but on groups of animals with high average accuracy.     

Genomic selection in American mink (Neovison vison) using a single-step genomic best linear unbiased prediction model for size and quality traits graded on live mink

Genomic selection relies on single-nucleotide polymorphisms (SNPs), which are often collected using medium-density SNP arrays. In mink, no such array is available; instead, genotyping by sequencing (GBS) can be used to generate marker information. Here, we evaluated the effect of genomic selection for mink using GBS. We compared the estimated breeding values (EBVs) from single-step genomic best linear unbiased prediction (SSGBLUP) models to the EBV from ordinary pedigree-based BLUP models. We analyzed seven size and quality traits from the live grading of brown mink. The phenotype data consisted of ~20,600 records for the seven traits from the mink born between 2013 and 2016. Genotype data included 2,103 mink born between 2010 and 2014, mostly breeding animals. In total, 28,336 SNP markers from 391 scaffolds were available for genomic prediction. The pedigree file included 29,212 mink. The predictive ability was assessed by the correlation (r) between progeny trait deviation (PTD) and EBV, and the regression of PTD on EBV, using 5-fold cross-validation. For each fold, one-fifth of animals born in 2014 formed the validation set. For all traits, the SSGBLUP model resulted in higher accuracies than the BLUP model. The average increase in accuracy was 15% (between 3% for fur clarity and 28% for body weight). For three traits (body weight, silky appearance of the under wool, and guard hair thickness), the difference in r between the two models was significant (P < 0.05). For all traits, the regression slopes of PTD on EBV from SSGBLUP models were closer to 1 than regression slopes from BLUP models, indicating SSGBLUP models resulted in less bias of EBV for selection candidates than the BLUP models. However, the regression coefficients did not differ significantly. In conclusion, the SSGBLUP model is superior to conventional BLUP model in the accurate selection of superior animals, and, thus, it would increase genetic gain in a selective breeding program. In addition, this study shows that GBS data work well in genomic prediction in mink, demonstrating the potential of GBS for genomic selection in livestock species.     

Effect of paratuberculosis on culling, milk production, and milk quality in dairy herds

OBJECTIVE: To determine the effect of paratuberculosis on culling, milk production, and milk quality in infected dairy herds. DESIGN: Cross-sectional study. ANIMALS: 689 lactating dairy cows in 9 herds. PROCEDURE: Milk, blood, and fecal samples were obtained from all cows. Fecal samples were evaluated via mycobacterial culture. Serum samples were tested with a commercially available ELISA for antibodies against Mycobacterium avium subsp paratuberculosis, and preserved milk samples were tested with an ELISA for antibodies against M paratuberculosis. Mixed effect and proportional hazards models were used to determine the effect of paratuberculosis on 305-day milk, fat, and protein production; somatic cell count linear score; and the risk of culling. RESULTS: Cows with positive results of bacteriologic culture of feces and milk ELISA produced less milk, fat, and protein, compared with herdmates with negative results. No difference in 305-day milk or fat production was detected in cows with positive results of serum ELISA, compared with seronegative cows. The 3 survival analyses revealed that cows with positive results of each test were at higher risk of being culled than cows with negative results. Paratuberculosis status, as determined by use of all 3 diagnostic tests, was not associated with milk somatic cell count linear score. CONCLUSIONS AND CLINICAL RELEVANCE: Results suggest that for the 9 herds in this study, paratuberculosis significantly decreased milk production and cow longevity.     

Factors associated with spread of pseudorabies virus among breeding swine in quarantined herds.

Knowledge of the factors that place susceptible gilts at highest risk of pseudorabies virus (PRV) infection in a quarantined herd is crucial to reduce spread of PRV within the herd. Cohorts of PRV seronegative gilts were monitored in 17 herds that were endemically infected with PRV to determine the location of breeding females at the time of infection with PRV and identify herd characteristics and management and housing factors that may influence spread of PRV in the breeding section of swine herds endemically infected with PRV. Blood samples were collected every 1 to 2 months for an average of 13.6 months. In addition, blood was collected from a representative sample of finishing pigs (greater than or equal to 20 weeks old) 3 times per year to determine their serologic PRV status. Incidence rates and relative risks of PRV infection were estimated for 4 areas of the breeding section: gestation barn, gilt pool, farrowing room, and breeding area. Overall, 28, 11, 8, and 2 females became infected with PRV in each of these areas, respectively. The greater number of females infected in the gestation barns, compared with the number of females infected in other locations, is probably a consequence of being at risk for a longer period rather than of a higher incidence rate. Herd size, common housing for gilts in the gilt pool and sows, and serologic pattern of PRV infection in finishing pigs were associated with the detection of spread of PRV in the breeding section of the 17 herds.(ABSTRACT TRUNCATED AT 250 WORDS)     

The impact of truncating data on the predictive ability for single‐step genomic best linear unbiased prediction

Simulated and swine industry data sets were utilized to assess the impact of removing older data on the predictive ability of selection candidate estimated breeding values (EBV) when using single‐step genomic best linear unbiased prediction (ssGBLUP). Simulated data included thirty replicates designed to mimic the structure of swine data sets. For the simulated data, varying amounts of data were truncated based on the number of ancestral generations back from the selection candidates. The swine data sets consisted of phenotypic and genotypic records for three traits across two breeds on animals born from 2003 to 2017. Phenotypes and genotypes were iteratively removed 1 year at a time based on the year an animal was born. For the swine data sets, correlations between corrected phenotypes (Cp) and EBV were used to evaluate the predictive ability on young animals born in 2016–2017. In the simulated data set, keeping data two generations back or greater resulted in no statistical difference (p‐value > 0.05) in the reduction in the true breeding value at generation 15 compared to utilizing all available data. Across swine data sets, removing phenotypes from animals born prior to 2011 resulted in a negligible or a slight numerical increase in the correlation between Cp and EBV. Truncating data is a method to alleviate computational issues without negatively impacting the predictive ability of selection candidate EBV.     

运用BLUP选种效果的探讨

应用多性状动物模型BLUP法，经四年选育湖北省畜牧良种场外种猪的产肉性能从表型和遗传上得到显著提高，长白猪、约克夏和杜洛克体重达100千克的背膘厚分别为1997年的16.5、19.7和19.1毫米下降到2000年的13.2、15.1和15.3毫米，体重达100千克日龄分别从1997年的183.5、188.7和194.3天下降到2000年的164.3、179.7和179.5天；背膘厚的年度平均育值分别由1995年的1.24、0.95和0.92甜言蜜语米下降到2000年的-0.48、-0.29和-0.21毫米，达100千克日龄的年均育种值分别为1995年3.84、0.84和5.69天下降到2000年的-0.38、-0.30和-0.55天，两性状的育种值指数分别从1995年的67.9、67.3和60.3点提高到2000年的110.5、106.2和105.9点。