Multibreed Model Research Articles

ESTIMATE OF GENETIC COMPONENTS OF BIRTH WEIGHT USING MULTI-BREED MODELS WITH DIFFERENT PEDIGREE STRUCTURES IN MESTIZO SHEEP

The objectives of this paper were to verify the influence of the multi-breed model and pedigree structure in the estimates of the genetic components to birth weight in a mestizo herd. Using 1234 birth weight records in two ways, one dataset with complete pedigree information (n = 1028) and another with incomplete pedigree information (n = 1234). The pedigree was composed for 10 sires, 366 and 448 dams for complete and incomplete pedigree, respectively. Used for analysis the maximum restricted likelihood method about the animal model, considering the birth weight trait, for a dataset with or not complete pedigree, used the WOMBAT program. The direct heritability coefficients in both datasets did not present large difference, 0.06 and 0.09 to best models in the complete and incomplete pedigree, respectively. However, in the incomplete pedigree, the maternal heritability was 0.29. The residual variance, which on the best model of the dataset with complete pedigree presented 0.51 and 0.35 for the best model of the dataset with incomplete pedigree. Concluding, we find the influence of de multi-breed model only in the dataset with complete pedigree, that considered the direct and breed effects. In the dataset with incomplete pedigree, the better model was that considered the direct and maternal additive effects.

Homoscedastic or heteroscedastic inference in the genetic evaluation of a multi-breed dairy cattle

The objective was to identify the best multi-breed model for the genetic evaluation of test-day milk yield (TDMY) in a Girolando population. The observations were measured from 2000 to 2014, totaling 122,507 observations of 13,544 first-lactation cows and a complete kinship matrix of 66,384 animals. Four different methodologies were evaluated: homoscedastic or heteroscedastic multi-breed model with Gaussian destruction (MAM-G-Ho or MAM-G-He) and homoscedastic or heteroscedastic multi-breed model with Student's t-distribution (MAM-T-Ho or MAM-T-He). The criteria for choosing the best model were the deviance information criterion (DIC), deviance based on the Conditional Predictive Ordinate (DCPO), and deviance based on the Bayes factors (DBF). According to the selection criteria, the MAM-T-He presented the lowest values of DIC, DCPO, and DBF, indicating a better quality of fit for the population in question. Heritability estimates were of medium magnitude for TDMY, varying from 0.27±0.02 to 0.39±0.03 in the different models. Spearman's correlations based on estimated breeding values used to compare the ranking of animals between Student's t and Gaussian models were low (0.63 to 0.72) when considering all sires or when considering 10, 20, or 30% of the best sires. These results support the implementation of robust models that account for sources of heteroscedasticity to increase the accuracy of genetic evaluations in the multi-breed population of the Girolando breed.

Genetic evaluation models for post-weaning weight gain in a multibreed Angus-Nelore population

Abstract: The objective of this work was to identify the most suitable model for the genetic evaluation of post-weaning weight gain in a multibreed Angus-Nelore population. Three models were tested using the Bayesian inference method: traditional animal model (M1), multibreed animal model without (M2) and with segregation (M3). The choice of the best model followed the criteria: number of parameters (Np), deviance information criterion (DIC), conditional predictive ordinate (CPO), and deviance based on Bayes factors. Spearman’s rank correlations were estimated for the top 10, 20, and 30% sires. M1 presented the highest values for all criteria, except for Np, and the lowest direct heritability estimate of 0.15±0.01. The heritability estimates for M2 and M3 were higher and similar, being 0.29±0.02 and 0.27±0.02, respectively. M3 showed the lowest values for mean deviance, DIC, and CPO, being the best-fitting model among the three tested. Spearman’s correlation between the predicted genetic values for the models ranged from 0.69 to 0.99. The multibreed models are the most suitable for the genetic evaluation of multibreed populations, and M3 shows the best fit for the studied population.

Usefulness of multi-breed models in genetic evaluation of direct and maternal calving ease in Holstein and Belgian Blue Walloon purebreds and crossbreds

The objective of this study was to verify the feasibility of a joint genetic evaluation system for calving ease trait of Belgian Blue (BBB) and Holstein (HOL) Walloon cattle based on data of purebred and crossbred animals. Variance components and derived genetic parameters for purebred BBB and HOL animals were estimated by using single-breed linear animal models. This analysis showed clear genetic differences between breeds. Estimates of direct and maternal heritabilities (±standard error) were 0.34 (±0.02) and 0.09 (±0.01) for BBB, respectively, but only 0.09 (±0.01) and 0.04 (±0.01) for HOL, respectively. Moreover, a significant negative genetic correlation between direct and maternal effects was obtained in both breeds: −0.46 (±0.04) for BBB and −0.29 (±0.11) for HOL. Variance components and derived genetic parameters for purebred BBB and HOL and crossbred BBB × HOL cattle were then estimated by using two multi-breed linear animal models: a multi-breed model based on a random regression test-day model (Model MBV), and a multi-breed model based on the random regression multi-breed model (Model MBSM). Both multi-breed models use different functions of breed proportions as random regressions, thereby enabling modelling different additive effects according to animal's breed composition. The main difference between these models is the way in which relationships between breeds are accounted for in the genetic (co)variance structure. Genetic parameters differed between single-breed and multi-breed analysis, but are similar to the literature. For BBB, estimates of direct and maternal heritabilities (±SE) were 0.45 (±0.07) and 0.08 (±0.01) by using Model MBV, and 0.45 (±0.08) and 0.09 (±0.02) for Model MBSM, respectively. For HOL, these estimates were 0.18 (±0.04) and 0.05 (±0.01) using Model MBV, and 0.16 (±0.04) and 0.05 (±0.01) for Model MBSM, respectively. Reliability gains (up to 25%) indicated that the use of crossbred data in the multi-breed models had a positive influence on the estimation of genetic merit of purebred animals. A slight re-ranking of purebred sires and maternal grandsires was observed between single-breed and multi-breed models. Moreover, both multi-breed models can be considered as quasi-equivalent models because they performed almost equally well with respect to MSE and correlations, for purebred and crossbred animals.

Genotype Environment interactions for milk production traits in Holstein and crossbred Holstein-Zebu cattle populations estimated by a character state multibreed model

A total of 478,555 lactation records from 203,825 cows calving from January 1984 to December 2003 were used to quantify the importance of the genotype-environment interaction (G×E) on milk production in a large population of Holstein and Holstein-Zebu crossbred animals under Cuban tropical conditions. These cows were distributed in a total of 1288 herds, daughters of 1565 Holstein (H) sires and 831 Holstein crossbred sires (HZ) with female Zebu (Z). Due to the dramatic political crisis in eastern countries at the end of the 1980s, all economical activities in Cuba were affected, and particularly the resources available for the dairy sector were marginalize; therefore two contrasted environmental periods of time were created: the first one was the calving from 1984 to 1990 and the second from 1991 to 2003. This division was made to estimate the magnitude of the G×E. Data were analyzed with different character state multi-breed sire models. Accordingly to the genetic correlation between the same traits across the time periods, the G×E was negative and highly significant but with more intensity in H sires with regard to HZ sires. A very important genetic variation was found in plasticity and a group of robust or generalist sires were identified showing a very good performance in the two time-period environmental conditions. The importance of the use of a Character State Multi-breed Model and the plasticity concept for breeding programs in the tropics are highly suggested.

Use of random regression model as an alternative for multibreed relationship matrix

A random regression model is presented as an approximation for multibreed variance model. The approximation is derived using the splitted multibreed model where the single breeding value is split to the breed specific and their segregation terms. The random regression model allows extending the multibreed information easily to genomic data models. We present the approach by a simple example.

Componentes de variância e parâmetros genéticos em uma população multirracial Nelore-Angus sob enfoque Bayesiano

Os objetivos neste estudo foram utilizar um modelo animal multirracial para verificar a presença de heterogeneidade de variância genética entre as raças Angus e Nelore, avaliar a importância da variância devida à segregação entre essas duas raças e estimar os efeitos genéticos aditivos de raça, dominância e epistasia aditiva × aditiva. Foram utilizados neste estudo 4.016 registros de ganho de peso pós-desmame do rebanho de formação da raça Brangus-Ibagé da Embrapa Pecuária Sul, analisados por três modelos distintos: modelo animal multirracial com segregação, modelo animal multirracial sem segregação; e modelo animal tradicional. A heterose reduziu substancialmente em cruzamentos avançados. A variância genética média a posteriori da raça Angus (189 ± 26 kg²) foi maior que o dobro da obtida para a raça Nelore (82 ± 23 kg²) e a variância da segregação entre essas raças foi de 11 ± 5 kg². A alteração das estimativas das variâncias genéticas raciais foi insignificante quando a variância da segregação foi fixada em zero. Critérios bayesianos de escolha apontam o modelo multirracial sem segregação entre raças como uma alternativa parcimoniosa adequada para avaliação genética dessa população.

Estimation of test-day model (co)variance components across breeds using New Zealand dairy cattle data

In New Zealand, a large proportion of cows are currently crossbreds, mostly Holstein-Friesians (HF) × Jersey (JE). The genetic evaluation system for milk yields is considering the same additive genetic effects for all breeds. The objective was to model different additive effects according to parental breeds to obtain first estimates of correlations among breed-specific effects and to study the usefulness of this type of random regression test-day model. Estimates of (co)variance components for purebred HF and JE cattle in purebred herds were computed by using a single-breed model. This analysis showed differences between the 2 breeds, with a greater variability in the HF breed. (Co)variance components for purebred HF and JE and crossbred HF × JE cattle were then estimated by using a complete multibreed model in which computations of complete across-breed (co)variances were simplified by correlating only eigenvectors for HF and JE random regressions of the same order as obtained from the single-breed analysis. Parameter estimates differed more strongly than expected between the single-breed and multibreed analyses, especially for JE. This could be due to differences between animals and management in purebred and nonpurebred herds. In addition, the model used only partially accounted for heterosis. The multibreed analysis showed additive genetic differences between the HF and JE breeds, expressed as genetic correlations of additive effects in both breeds, especially in linear and quadratic Legendre polynomials (respectively, 0.807 and 0.604). The differences were small for overall milk production (0.926). Results showed that permanent environmental lactation curves were highly correlated across breeds; however, intraherd lactation curves were also affected by the breed-environment interaction. This result may indicate the existence of breed-specific competition effects that vary through the different lactation stages. In conclusion, a multibreed model similar to the one presented could optimally use the environmental and genetic parameters and provide breed-dependent additive breeding values. This model could also be a useful tool to evaluate crossbred dairy cattle populations like those in New Zealand. However, a routine evaluation would still require the development of an improved methodology. It would also be computationally very challenging because of the simultaneous presence of a large number of breeds.

Multi‐breed genetic evaluation in a Gelbvieh population

A multi-breed model was presented for the genetic evaluation of growth traits in beef cattle. In addition to the fixed effects, random direct and maternal genetic effects, and random maternal permanent environmental effects are considered; the model also fits direct and maternal heterosis and direct and maternal breed-of-founder (BOF) x generation group effects using a Bayesian approach that weights prior literature estimates relative to information supplied by the dataset to which the model will be applied. The multi-breed evaluation procedures also allow the inclusion of external evaluations for animals of other breeds. The multi-breed model was applied to a dataset provided by the American Gelbvieh Association. Different analyses were conducted by varying the weights given to the prior literature relative to the information provided by the dataset. Large differences were observed for the heterosis estimates, the BOF x generation group effect estimates, and the predicted breeding values across breeds due to the weights posed on prior literature estimates versus estimates derived directly from data. However, the rankings within breed were observed to be relatively robust to the different weights on prior information.

Genetic Evaluations for Mixed-Breed Populations

An all-breed animal model was developed for routine genetic evaluations of US dairy cattle. Data sets from individual breeds were combined, and records from crossbred cows were included. About 1% of recent cows were first-generation crossbreds. The numbers of cows with records since 1960 ranged from 10 to 22million for the 6 traits analyzed, which were milk, fat, protein, somatic cell score, productive life, and daughter pregnancy rate. Programs were modified to account for general heterosis, to group unknown parents separately by breed, to adjust variances separately by breed, and to adjust data to a 36-mo age equivalent instead of a mature equivalent. Convergence rate of the all-breed model was similar to that of the previous within-breed animal model. Estimated breed differences were similar to those obtained previously from phenotypic breed means or from studies of crossbred cows and their herd-mates. Genetic evaluations from the all-breed and within-breed systems had high correlations: >0.99 for recent Holsteins and slightly <0.99 for other breeds. Predicted transmitting abilities will be converted back to the within-breed bases for purebred animals and to the breed of sire base for crossbred animals so that most purebred breeders will not be affected by the change to a multibreed model. Evaluations of crossbred animals from the multibreed model can include accurate information for both parents. Reliabilities also increase for purebred relatives because of the additional crossbred records and in mixed breed herds because cows of other breeds are additional contemporaries. Another benefit of the multibreed model is that breed differences are routinely estimated and updated. More research and education may be needed on using the new evaluations in the design of breeding programs. Implementation is expected in May 2007.