动物育种中的方差组分分析及统计软件包

方差组分估测是分析变异来源和预测育种值的基础,国内外对其进行了大量的研究,对动物育种中方差组分的估测方法及统计分析软件包作了一个简单的综述.     

甘肃高山细毛羊数量性状遗传相关和通径分析

为了探讨甘肃高山细毛羊数量性状的遗传参数,研究应用2006—2007年甘肃皇城绵羊繁育技术推广站采集的甘肃高山细毛羊6个核心群育种资料,对初生重,出生等级,断奶毛长,断奶重,1.5岁毛长、细度、产毛量及体重的遗传力以半同胞组内相关法进行估测,对遗传相关以公羊的方差组分和协方差组分进行估测,并对计算出的相关系数进行显著性检验。结果表明:以上性状对产毛性状的影响存在很大差异,通径分析进一步揭示污毛量、净毛率及与其互作的性状(断奶重、剪毛后体重、毛丛长度、束强)是影响产毛量最主要的因素。     

复合育种值电算估测

种畜的复合育种值估测一直是从事家畜育种工作的同志比较关注的问题。复合育种值估测是指将同一性状的各种亲属(或包括种畜本身)的记录资料合并估计的育种值。由于复合育种值具有多种亲属的遗传信息,故根据它来选种要比仅根据单项资料估测出的育种值来选种效果更准确可靠。然而由于复合育种值估测较为复杂,读者可参见吴仲贤教授著《统计遗传学》(1977年科学出版社出版)。该书针对当时我国多数生产单位尚未有电算工具的情况,列举了各种资料不同组合的偏回归系数计算公式达200多个,其中多数公式的计算都比较繁琐。     

遗传学应用于畜牧生产第三次世界大会

本次大会于1986年7月在美国内布拉斯加大学召开。95人为大会正式发言者(其中70人来自美国以外),另有200篇书面发表的论文。大会主题为:育种目标、种畜的利用和未来的育种方案、分子遗传学和家畜改良、生长组分的遗传学、食欲和饲料利用、抗病力和环境适应性、繁殖技术、选育实验、参数估测和遗传资源的管理。     

对蛋鸡育种的探讨和一种估测产蛋重的新方法

我国商品蛋主要以重量论价,而育种场主要根据产蛋数选种。本文估测了京白Ⅱ系近交系4个世代若干数量性状的表型和遗传参数,并一律采用入舍鸡产量。结果表明,如果改变我国蛋鸡育种的以往方法,直接对产蛋重选种,可望带来经济效益上的真正进展。本文还用Г函数配合了日产蛋重曲线,并介绍了查表积分估测法,为估计产蛋重提供了一种新方法。     

试论遗传参数的估测

从1979年开始,三年来动物数量遗传理论及其应用科研协作组在农业部科技局和畜牧兽医总局科技处的领导和支持下,直接组织了十多个协作单位,还带动了不少其他单位,对我国境内的二十多个畜禽品种的部分性状,进行了遗传参数的估测工作。这次估测工作的规模之大,在我国动物遗传和畜禽育种史上是空前的。大家的辛勤劳动带来了很有价值的丰硕成果,为我国畜禽育种工作提供了一批本国畜群的遗传参数。这是我国畜禽育种工作的一项非常重要的基本建设工     

应用BLUP方法选择秦川公牛的研究

以最佳线性无偏预测法(BIUP:Best Linear unbiased prediction)估测了13头秦川种公牛的初生重、断奶重、日增重、体高、胸围、体斜长6性状的育种值和复合育种值。用约束最大似然估计法(REML:Restricted maximum likelihood)估计方差组分,得到以上6性状的遗传力分别为0.24,0.215,0.29,0.48和0.36。由公牛加权平均传递力(WATA,weight average transmitting ability)对时间序列的回归估计群体的年平均遗传进展分别为0.1859,0.0436,1.74(kg/yr)和0.2304,0.231,0.1761(cm/yr)。分析了以往种公牛利用的合理性问题。研究表明用BIUP法选择秦川公牛是适用和有效的。     

东北梅花鹿产茸能力估测方法的研究

本研究给出了估测、比较具有不同次产茸记录公鹿产茸能力的方法,为科学地评价公鹿的产茸能力提供了依据,对茸鹿育种具有指导意义。     

伊莎明星肉鸡母系主要经济性状遗传参数的估计

本文围绕明星肉鸡的选育问题,对母系主要经济性状的遗传参数进行了估测,探讨了各系的遗传特性和各性状间的相互关系,为今后优化育种提供了理论依据。     

BLUP和PD法在小群混合世代种牛育种值测定的对比

本文用10头西门塔尔公牛168个女儿的材料对比同期同龄法和布拉普(BLUP)法来估测育种值。两种方法的结果基本一致。在小样本的情况下布拉普法可验证同期同龄法的可靠性。引入亲缘矩阵和世代矩阵在混合世代的情况下,不以年度的表型值的进展为转移,可以估测各代的各自遗传效应。BLUP法可对无后裔的公牛进行育种值的预报。     

Estimation of genetic parameters on test stations using purebred and crossbred progeny of sires of the Bavarian Piétrain

Genetic parameters were estimated for purebred and crossbred progeny of Bavarian Piétrain sires on two test stations. The data set used contained 4276 purebred pigs and 13,980 crossbred pigs recorded between 2000 and 2004. In total 332 sires having purebred and crossbred progeny were available to estimate the genetic correlations between purebred and crossbred performances. Though the genetic correlations between purebred and crossbred pigs are fairly high (0.7–0.9), their performances have to be considered as genetically different traits, because variance components and heritabilities differ substantially. Therefore, purebred and crossbred breeding values of candidates are not identical, and thus result in different rankings. However, due to the high correlations purebred pigs provide a lot of information for estimating the crossbred breeding values of the real selection criterion. The Halothan locus, whose effects have been analyzed in detail, affects both purebred and crossbred parameters. To avoid detrimental effects on the efficiency of the breeding programme, the n-allele could be either eliminated or the genotypes of all test animals should be known. Differences in the variance components between the two test stations have been found and are problematic with respect to the breeding value estimation utilizing the pooled data set. Hence, it should be attempted to further improve the standardization of the performance test on both stations.     

Technical note: prediction of breeding values using marker-derived relationship matrices

In livestock populations, estimation of breeding values for selection requires a matrix describing the additive relationship between individuals in the population. This matrix can be derived from pedigree information. In some livestock populations, pedigree information may be unavailable, incomplete, or in error. Here we use simulated data to demonstrate that marker-derived relationship matrices can be used to predict breeding values and estimate additive variance components, provided the markers are sufficiently dense. The approach is demonstrated for an Angus data set with 9,323 SNP markers genotyped.     

Estimation of additive and non-additive genetic effects for fertility and reproduction traits in North American Holstein cattle using genomic information

Non-additive genetic effects are usually ignored in animal breeding programs due to data structure (e.g., incomplete pedigree), computational limitations and over-parameterization of the models. However, non-additive genetic effects may play an important role in the expression of complex traits in livestock species, such as fertility and reproduction traits. In this study, components of genetic variance for additive and non-additive genetic effects were estimated for a variety of fertility and reproduction traits in Holstein cattle using pedigree and genomic relationship matrices. Four linear models were used: (a) an additive genetic model; (b) a model including both additive and epistatic (additive by additive) genetic effects; (c) a model including both additive and dominance effects; and (d) a full model including additive, epistatic and dominance genetic effects. Nine fertility and reproduction traits were analysed, and models were run separately for heifers (N = 5,825) and cows (N = 6,090). For some traits, a larger proportion of phenotypic variance was explained by non-additive genetic effects compared with additive effects, indicating that epistasis, dominance or a combination thereof is of great importance. Epistatic genetic effects contributed more to the total phenotypic variance than dominance genetic effects. Although these models varied considerably in the partitioning of the components of genetic variance, the models including a non-additive genetic effect did not show a clear advantage over the additive model based on the Akaike information criterion. The partitioning of variance components resulted in a re-ranking of cows based solely on the cows’ additive genetic effects between models, indicating that adjusting for non-additive genetic effects could affect selection decisions made in dairy cattle breeding programs. These results suggest that non-additive genetic effects play an important role in some fertility and reproduction traits in Holstein cattle.     

Mate selection strategies to exploit across- and within-breed dominance variation

In multi‐breed livestock populations, dominance variance is found both between and within breeds for many economically important traits. Mate selection strategies were developed to exploit both types of dominance variation simultaneously, with the aim of maximizing genetic merit of progeny. The extended super‐breed model, in which breeds are viewed as groups of related and inbred animals within a ‘super‐breed’, was used to predict individual additive and dominance effects for use in mate selection. Performance of mate selection was assessed under a range of relative values of additive and dominance variances for one generation of breeding. Mate selection on total progeny merit, including additive effects, individual dominance effects, and value of heterosis, was the optimal breeding strategy at all values of (co)variance components, with improvements in total progeny performance of up to 12.5 % over truncation selection followed by random mating when dominance variance was large relative to total genetic variance. Improvement in progeny merit from mate selection, relative to truncation selection, followed by random mating or truncation selection, followed by mate allocation, was particularly great (up to 53 %) when there was considerable heterosis. Improvements were small if dominance variance was small relative to total genetic variance, and heterosis was low. If the target population is large, full mate selection on total progeny merit is computationally demanding, and unlikely to be practical. Alternative, less computationally demanding strategies made nearly optimal selection and mating decisions at some parameter estimates. Integrating multi‐breed genetic evaluation, using a superbreed model, with mate selection provides a powerful framework for the design of breeding programmes which exploit available sources of genetic merit.     

Three statistical approaches for genetic analysis of disease resistance to vibriosis in Atlantic cod (Gadus morhua L.)

The objective of this study was to estimate and compare variance components and sire breeding values for disease resistance to vibriosis in Atlantic cod (Gadus morhua L.) using 3 statistical approaches. A total of 3,576 individually tagged juvenile cod from 50 full-sib families were infected with Vibrio anguillarum, which causes vibriosis, a frequently reported disease in cod aquaculture. The experimental fish were progeny of captured wild cod from populations of southern coastal cod (POP1), and northern coastal cod and northeast Arctic cod (combined as POP2 in the genetic analyses). Fish were randomly assigned to 1 of 3 test tanks, and daily mortality was recorded until the termination of the experiment at d 31 postinfection. Variance components were estimated separately for the 2 populations using a Cox regression model, univariate linear model, and a linear model that accounted for censoring. With all approaches, the additive genetic sire variance estimated from POP1 was greater than for POP2. Heritability estimates across populations varied from 0.08 to 0.17 depending on the method used. The Cox regression model and univariate linear model resulted in greater heritability estimates for POP1 (0.10 and 0.16) than for POP2 (0.08 and 0.13), whereas the contrary was true with a linear model that accounted for censoring (0.17 vs. 0.14). The predicted breeding values for the sires from the 3 approaches were highly correlated (0.97 to 0.99). This is likely due to the fact that censoring only occurred at the end of the test; i.e., observations of the most resistant fish were censored. The considerable genetic variation found in this study suggests that vibriosis resistance may be improved through selective breeding. The univariate linear model, even without censoring of the data, was robust for the estimation of breeding values using the present data. Therefore, inclusion of vibriosis resistance in the multivariate linear estimation of breeding values for the traits of economic importance in Atlantic cod seems appropriate.     

Empirical Percentile Growth Curves with Z-scores Considering Seasonal Compensatory Growths for Japanese Thoroughbred Horses

Percentile growth curves are often used as a clinical indicator to evaluate variations of children’s growth status. In this study, we propose empirical percentile growth curves using Z-scores adapted for Japanese Thoroughbred horses, with considerations of the seasonal compensatory growth that is a typical characteristic of seasonal breeding animals. We previously developed new growth curve equations for Japanese Thoroughbreds adjusting for compensatory growth. Individual horses and residual effects were included as random effects in the growth curve equation model and their variance components were estimated. Based on the Z-scores of the estimated variance components, empirical percentile growth curves were constructed. A total of 5,594 and 5,680 body weight and age measurements of male and female Thoroughbreds, respectively, and 3,770 withers height and age measurements were used in the analyses. The developed empirical percentile growth curves using Z-scores are computationally feasible and useful for monitoring individual growth parameters of body weight and withers height of young Thoroughbred horses, especially during compensatory growth periods.     

Genetic parameters for within-litter variation in piglet birth weight and change in within-litter variation during suckling

The objective of this study was to ascertain whether maternal additive genetic variance exists for within-litter variation in birth weight and for change in within-litter variation in piglet weight during suckling. A further objective was to estimate maternal genetic correlations of these two traits with mortality, birth weight, growth, and number of piglets born alive. Data were obtained from L?vsta research station, Swedish University of Agricultural Sciences, and included 22,521 piglets born in 2,003 litters by 1,074 Swedish Yorkshire sows. No cross fostering was used in the herd. The following seven traits were analysed in a multivariate animal (sow) model: number of piglets born alive, within-litter SD in birth weight, within-litter SD in piglet weight at 3 wk of age, mean weight at birth, mean weight at 3 wk of age, proportion of stillborn piglets, and proportion of dead piglets during suckling. Maternal genetic variance for the change in within-litter SD in piglet weight during suckling was assessed from the estimated additive genetic covariance components by conditioning on within-litter SD in birth weight. Similarly, mean growth of piglets during suckling was assessed from the additive genetic covariance components by conditioning on mean weight at birth. The heritability for within-litter SD in birth weight was 0.08 and 0.06 for within-litter SD in piglet weight at 3 wk. The genetic correlation between these two traits was 0.71. Little maternal genetic variance was found for the change in within-litter SD in piglet weight during suckling, and opportunity for genetic improvement of this trait by selective breeding seems limited. The genetic correlation of within-litter SD in birth weight with proportion of dead piglets during suckling was 0.25 and of within-litter SD in birth weight with mean growth of piglets was -0.31. The maternal genetic variance and heritability found for within-litter SD in birth weight indicates that genetic improvement of this trait by selective breeding is possible. In addition, selection for sows' capacity to give birth to homogeneous litters may be advantageous for piglet survival, piglet growth, and litter homogeneity at weaning.     

Impact of dominance effects on sow longevity

The purpose of the current study was to estimate variance components, especially dominance genetic variation, for overall leg action, length of productive life and sow stayability until third and fifth parity in the Finnish pig populations. The variance components were estimated in two purebred [Landrace (LR), n = 23 602 and Large White (LW), n =22 984] and crossbred (LR × LW, n = 17 440) data sets. Five different analyses were carried out for all the traits to compare the effect of sows’ inbreeding, common litter environment and parental dominance in the statistical model when determining the genetic correlations of the traits for the two purebred and crossbred populations. Estimated heritabilities for the traits ranged from 0.04 to 0.06. The estimates for the proportion of dominance variance of phenotypic variance (d²) varied between 0.01 and 0.17, and was highest in the crossbred dataset. The genetic correlations of the same traits in purebred and crossbred were all high (>0.75). Based on current results, the effect of dominance should be accounted for in the breeding value estimation of sow longevity, especially when data from crossbred animals are included in the analyses. Because dominance genetic variation for sow longevity exists that variation should be utilized through planned matings in producing sows for commercial production.     

Estimation of variance components and prediction of breeding values using pooled data

Data on performance of animals are, in several situations, collected at the group rather than individual level. Genetic evaluations in farm animals, however, are based on phenotypic information collected at the individual level. Therefore, it would be very attractive to extend genetic evaluations by incorporating information collected at the group level. In this paper we show the use of data collected at the group level for the estimation of variance components and the prediction of breeding values. We outline a general procedure that can be applied in different farm animal species. In the present work this procedure was applied to BW, for which pooled, as well as individual, observations were available, thus allowing for a comparison of the estimates, and to egg production, for which only pooled data were available. For BW at 19 and 27 wk the estimated heritabilities based on individual observations were very similar to those based on pooled observations. For BW at 43 and 51 wk, heritability estimates based on individual and pooled data were different, which can be caused by the emergence of competition effects. The accuracy of EBV predicted from pooled observations was about 70 to 80% of the accuracy of EBV predicted from individual observations. This result quantifies the loss deriving from the use of pooled instead of individual observations. Results show that estimation of variance components and breeding values from pooled data instead of individual observations is theoretically and practically feasible.     

Reduced animal model with differential genetic grouping for direct and maternal effects.

Mixed-model equations for the reduced animal model with maternal effects and different genetic grouping of unknown parents for additive direct and maternal effects are derived. The matrices that relate the expected value and the variance of the breeding values of non-parents to the parents, as well as the different contributions of parental and non-parental breeding values, to the resulting mixed-model equations are presented. Mis-specification of additive maternal variance and the additive covariance between direct and maternal effects, arising from missing information on the dams of known individuals with records, is discussed. To avoid an incorrect specification of the variance-covariance matrix of the records without having to invert a nondiagonal variance of the residual terms, the breeding values of the unknown dams of individuals with records are included in the equations. Breeding values of non-parents are back-solved after the solutions for genetic groups and breeding values of parents are computed as simply as in cases in which maternal effects are absent. A numerical example is included to illustrate the derivations.