动态性状育种值的估计(预测)方法探讨

为探讨更具一般性的动态性状育种值估计方法,在二步估计法的基础上建立了用于动态性状育种值估计(预测)的数学模型,采用该模型可实现对动态性状育种值的一步估计(预测),并保证估计结果的最优性。按线性与非线性模型两种情况分别论述了动态性状育种值的估计(预测)方法。结果表明:与随机回归模型相比,所提出的数学模型不仅具有明显的优越性,更具一般性;还证明了当每个个体动态性状的观测点一一对应且个体之间不存在亲缘关系时,一步分析法与二步分析法的估计结果相同。     

利用广义线性方法定位家畜抗性等级性状的QTL

在广义线性模型的框架内模拟研究了家畜抗性等级性状的QTL定位方法，QTL参数的估计采用最大似然方法，比较了阈模型方法与一般线性方法的QTL定位效率，并对影响等级性状QTL定位效率的主要因素（QTL效应、性状的遗传力）进行了模拟研究。试验为多个家系的女儿设计，资源群体大小为500头。结果表明：在QTL位置参数估计及检验功效方面，阈模型方法具有一定的优势，对抗性等级性状QTL定位的功效也高于线性方法。另外，性状遗传力和QTL效应的大小对QTL定位的准确度也有直接的影响，随着性状遗传力QTL效应的增大，2种方法QTL定位的效率均有不同程度的提高。     

性别对杜洛克猪主要经济性状的遗传参数估计的影响

通过建立不同模型，采用MTDFREML程序，对韩国京畿道二川郡的某原种猪场的3443头鉴定的杜洛克猪的资料进行了性别对杜洛克猪主要经济性状的遗传参数的研究。结果表明，杜洛克猪主要经济性状的最小二乘均值在公母猪之间有明显的差异（P＜0．05），模型中同时考虑母体效应时得到的遗传力均低于没有考虑母体效应时的遗传力。同时，将性别作为固定效应时的各性状间的遗传相关高于只考虑公猪时的各性状间的遗传相关，但低于只考虑母猪时的各性状间的遗传相关。     

双列杂交试验的多元分析

动物双列杂交试验中所用亲本品种的纯合性常相差很大，所以各遗传组的性状间误差协方差矩阵常不同。本文根据线性模型技术，提出了分析双列杂交试验的新方法，可处理多性状、多环境单元，给出了任一遗传组的综合杂种优势、遗传经济价值和生产力等参数的最优线性无偏估计和置信区间。     

母猪产仔模型的选优与模型参数的遗传分析

根据姜曲海猪产仔数表现出的生命旋回与逐胎波动两大特征，分别用泊松旋回模型、三次多项式模型、改进二次多项式模型Ⅰ和改进二次多项式模型Ⅱ4种数学模型对其窝产仔记录进行了拟合。从统计和生物意义两方面综合评价，改进二次多项式模型Ⅰ为产仔数的优选模型。单性状动物模型和DFREML算法被用来估计优选模型参数的遗传力。研究结果表明，产仔模型参数A、B和模型纯二次曲线顶点估计值的遗传力较产仔数的遗传力高，提示针对产仔模型参数的选择将比直接选择产仔数更为有效。     

亲嘉系猪选育效果的灰色预测

选用新嘉系猪０－５世代的育种材料，利用ＧＭ（１，１）模型对６月龄体重，活体膘厚、臀腿围、体长指数和瘦肉率等性状的选育效果进行动态模拟和预测，结果表明，至６世代时，各主要经济性状值都将达到或超过原定的育种目标。     

新嘉系猪选育效果的灰色预测

选用新嘉系猪０—５世代的育种材料，利用ＧＭ（１，ｌ）模型对６月龄体重、活体膘厚、臀腿围、体长指数和瘦肉率等性状的选育效果进行动态模拟和预测。结果表明，至６世代时，各主要经济性状值都将达到或超过原定的育种目标。     

母猪生产力性状的遗传分析

对二花脸猪、大白猪及其杂交一代3个群体共1686窝的生产力性状资料进行了遗传分析。用公畜模型估计了初生窝重、初生仔猪平均重、20日龄窝重、20日龄仔猪平均重、45日龄窝重、45日龄仔猪平均重的表型参数和遗传参数。与大白猪相比，二花脸猪抚育能力较强，生长速度较慢。前4个性状受母体影响较大，后2个性状受母体影响较小。6个性状的遗传力分别为0．251、0．277、0．341、0．330、0．235、0．186，表明直接选择有一定效果。从6个性状中选出初生窝重和20日龄窝重来反映母猪生产力，前者反映母猪繁殖力，后者反映母猪抚育能力。     

规模化猪场 猪群性能测定与种猪选择方案探讨

论述了规模化养猪性能测定的目的、内容与方法,阐述了种猪选留顺序、主选性状、选择强度、世代间隔等选种选配繁育制度,并提出了选种的综合性状参数法计算模型。     

浅议甘肃高山细毛羊重要经济性状的选择

为总结甘肃高山细毛羊选种工作经验，从经济性状的选择入手挖掘提高品种质量的新途径，在已获得的１８个性状遗传参数的基础上，对甘细羊重要经济性状１６年实际选择效果进行了测算，利用有关遗传变异参数对未来期望遗传进展进行了预测，对加速育种进程的方法进行了探讨     

长白和大白猪主要生长性状的遗传参数估计

旨在分析母猪的出生年份、出生季节、初生重、开测日龄等固定效应对长白、大白猪主要生长性状的影响,并对目标生长性状进行遗传参数估计（遗传力、遗传方差、表型相关和遗传相关）,为猪的遗传改良提供基本依据。本试验利用GLM模型分析试验猪群（398头长白猪和1 176头大白猪）的固定效应对猪生长性状的影响,并采用多性状动物模型对目标性状进行遗传参数估计。目标生长性状包括达100 kg体重日龄（age to 100 kg,AGE）、达100 kg背膘厚（backfat to 100 kg,BF）、100 kg平均日增重（average daily gain to 100 kg,ADG）。研究表明,在大白和长白猪中,猪的出生年、出生季、初生重以及开测日龄对生长性状均具有极显著的影响（P<0.001）;长白猪的AGE、ADG和BF的遗传力分别为0.321、0.327和0.324,大白猪对应性状的遗传力分别为0.454、0.469和0.408;长白猪的ADG和AGE之间的遗传相关、表型相关分别为-0.990、-0.995,大白猪的ADG和AGE之间的遗传相关、表型相关分别为-0.993、-0.998,均呈现较强的负相关。长白、大白猪的生长性状（AGE、ADG、BF）均属于中等遗传力性状,其出生年份、出生季节、初生重和开测日龄对猪的生长性状影响较大。在遗传参数估计分析时,提高样本数量并提升表型数据质量,可以增加遗传参数估计的可靠性。本研究中的生长性状遗传参数估计结果较为可靠,可为后续的遗传改良提供参考。     

Effects of integrated genetic evaluations for Icelandic horses on predictive ability, accuracy and selection bias

The genetic evaluation of Icelandic horses is currently based on results from breeding field tests of riding ability and conformation. The effect of integrating competition traits and/or test status into the genetic evaluation was studied concerning estimation bias, predictive ability, accuracy, correlations between breeding values and ranking of sires. Breeding field test data included 19 954 records from horses assessed in 11 countries during 1994-2008. Competition data included 44 160 records from 7687 horses competing in Iceland and Sweden in 1998-2008. Test status was defined as attendance of horses born in Iceland at breeding field tests and/or in competition. Overall, there were trivial differences between different genetic evaluation models regarding estimation bias and predictive ability. Very strong correlations were estimated between breeding values for combined indexes of conformation, riding ability and total score from different models. Higher accuracy was achieved for most of the traits when competition traits and/or test status were added to the model. Sires ranked differently when the new traits were added to the genetic evaluation model. It was concluded that competition traits should be integrated into the genetic evaluation. Further analyses on genetic parameters for test status and its relationship with the other traits are needed for future inclusion of test status in the genetic evaluation.     

Genotype by environment interaction and model comparison for growth traits of Santa Ines sheep

The objectives of the present study were to compare alternative models for the genetic evaluation and assess the importance of genotype by environment interaction (G×E) in the estimation of genetic parameters and genetic evaluation of birth weight (BW), weight at 60 days of age (W60) and weight at 180 days of age (W180) of Santa Ines sheep. Data comprise 7622 BW, 4673 W60 and 2830 W180 records from animals born in 44 Brazilian herds. Four models were used for the analyses: animal model (AM) with homogeneous residual variance (1), or heterogeneous residual variance (2), hierarchical reaction norms model (HRNM) with homogeneous (1) or heterogeneous residual variance (2). The models that best fit the BW, W60 and W180 data were AM2, HRNM1 and HRNM2 respectively. Thus, models for genetic evaluation that consider heterogeneity of variances are recommended to evaluate growth traits of sheep. The correlation between intercept and slope of the HRNM was higher than 0.70 for all traits studied, indicating that animals with higher average breeding values responded better to improvement in environmental conditions, a fact characterizing the scale effect of G×E. Therefore, G×E is an important factor to be considered in the estimation of genetic parameters and genetic evaluation of growth traits of sheep.     

“Bending” and beyond: Better estimates of quantitative genetic parameters?

Multivariate estimation of genetic parameters involving more than a handful of traits can be afflicted by problems arising through substantial sampling variation. We present a review of underlying causes and proposals to improve estimates, focusing on linear mixed model‐based estimation via restricted maximum likelihood (REML). Both full multivariate analyses and pooling of results from overlapping subsets of traits are considered. It is suggested to impose a penalty on the likelihood designed to reduce sampling variances at the expense of a little additional bias. Simulation results are discussed which demonstrate that this can yield REML estimates that are on average closer to the population values than their unpenalized counterparts. Suitable penalties can be obtained based on assumed prior distributions of selected parameters. Necessary choices of penalty functions and of the stringency of penalization are examined. We argue that scale‐free penalty functions lend themselves to a simple scheme imposing a mild, default penalty which can yield “better” estimates without being likely to incur detrimental effects.     

Genetic analyses of piglet survival and individual birth weight on first generation data of a selection experiment for piglet survival under outdoor conditions

Genetic parameters of piglet survival traits and birth weight were estimated on the first generation data of a selection experiment aimed at improving piglet survival using a multiple trait linear and threshold model. Data on 5293 piglets for survival at birth, at day one after birth and during the entire nursing period, as well as individual birth weight and litter size, were recorded in an outdoor production system. Genetic effects of piglet survival traits and birth weight were estimated based on threshold and Gaussian models, respectively, using a Bayesian approach. The statistical model included as fixed effects selection group, parity, gender, fostering, gestation length and month of farrowing and, alternatively, an adjustment for litter size. Direct genetic effects (i.e. the piglet's genetic potential) for piglet survival and birth weight were estimated separately, whereas maternal genetic and environmental effects could only be estimated for the given data structure in a combined litter effect. Posterior means of heritabilities for direct genetic effects of survival at birth, at first day after birth and the entire nursing period, as well as birth weight, were 0.08, 0.07, 0.08 and 0.20, respectively. Genetic correlations among survival traits were in the range of 0.29 to 0.40 and indicate that these traits were mainly attributable to different genetic effects. Genetic correlations between direct effects of survival traits and birth weight ranged between 0.18 and 0.23 and were reduced when weights of stillborn piglets were omitted in the analysis or the traits were adjusted for litter size. The magnitudes of direct genetic effects of survival traits are substantially higher than estimates in the literature, which may indicate that these traits have a higher genetic influence under outdoor conditions. The use of birth weight in the multiple trait estimation provided important information for the estimation of survival traits due to its favourable genetic correlations with survival, its high heritability and its high information content as a continuously measured trait.     

基于系谱和基因组信息估计荷斯坦青年母牛体重性状遗传参数

旨在设计利用不同信息来源的模型估计荷斯坦后备牛不同月龄体重性状的遗传参数。本研究于2014—2020年测定并收集了7 122头荷斯坦牛32 338条0～12月龄体重数据,分别利用系谱信息(linear mixed model with pedigree relationship matrix, LM＿A)和系谱-基因组信息构建亲缘关系矩阵(linear mixed model with genotype-pedigree joint relationship matrix, LM＿H),基于母体效应动物模型估计初生重,基于是否考虑初生重作为协变量的单性状动物模型估计2～12月龄各月龄体重遗传力,并利用双性状动物模型估计初生重与其它月龄体重的遗传相关。结果显示,对于初生重,根据赤池信息量准则(Akaike information criterion, AIC),LM＿H方法的拟合程度显著优于LM＿A方法,但两种方法估计的遗传参数相差不大：直接遗传力分别为0.30和0.32,母体遗传力分别为0.08和0.09,个体直接遗传效应和母体遗传效应遗传相关系数分别为-0.65和-0.64;对于2～...     

Multi-trait animal model estimation of genetic parameters for linear type and gait traits in the Belgian warmblood horse

Genetic parameters for the height at withers, 27 linear type and six linear gait traits were estimated for the Belgian warmblood horse. Observations on 987 mares, mostly 3 years old, were analysed using a multi-trait animal model. The statistical model included appraiser, age and location (date × place of appraisal) as fixed effects. Genetic parameters were estimated using a canonical transformation and an expectation-maximization restricted maximum likelihood algorithm with an additional deceleration step. Estimates of heritability for the 33 linear traits were between 0.15 and 0.55. Heritability of the height at withers was 0.34 ± 0.06. Estimated genetic correlations ranged from −0.60 to 0.98 with an average SE of 0.10. The highest positive correlations were found among traits of walk and among traits of trot. Volume and the quality of legs were the most negatively correlated. Estimated genetic parameters indicated that the linear scoring system is a valuable tool to assess conformation. The full (co)variance matrix is now available for breeding value estimation to support selection for conformation and gaits.     

Genetic parameters for litter traits at different parities in purebred Landrace and Yorkshire pigs

Comparison of the multi‐trait animal model and the traditional repeatability model was carried out using data obtained from 6,424 Landrace and 20,835 Yorkshire sows farrowed from January 2000 to April 2018 in order to estimate genetic parameters for litter traits at different parities. Specifically, records of the total number born (TNB), number born alive (NBA), total number of mortality (MORT), number of stillborn (NSB) and number of mummified pigs (MUM) were used. Although results showed the heterogeneity of heritability for litter traits at different parities, the mean heritability estimates from the multi‐trait model were found to be higher than those of the repeatability model for all traits in both pig breeds. In terms of genetic correlation between parities, a slight difference in genetic control in the first parity was noted for TNB and NBA in Landrace and Yorkshire pigs. The correlation between the first parity and later parities ranged from 0.48 to 0.74 for TNB and NBA in both breeds. Moreover, genetic correlation between parities for MORT and NSB was observed to be high for parities higher than 2 in Yorkshire pigs. For MUM, genetic correlation between the first and other parities was generally low in both breeds, indicating that culling pigs on the basis of MUM at the first parity could probably be unreasonable. Overall, the results of this study suggest that the multi‐trait approach for litter size traits is useful for the accurate estimation of genetic parameters.     

性别对约克夏猪的主要经济性状遗传参数估计的影响

本文采用动物模型DFREML ,对位于韩国京畿道二川郡的某原种猪场的 6 16 8头鉴定的约克夏猪的资料进行了性别对约克夏猪主要经济性状的遗传参数影响的研究。结果表明 :约克夏猪主要经济性状的最小二乘均值在公母猪之间有明显差异 (P <0 .0 5 ) ;模型中同时考虑母体效应时得到的遗传力均低于没有考虑母体效应时的遗传力。同时 ,将性别视为固定效应时各性状间的遗传相关高于只考虑公猪时各性状间的遗传相关 ,但低于只考虑母猪时各性状间的遗传相关