Breeding Values For Carcass Traits Research Articles

Prediction of Hanwoo Cattle Phenotypes from Genotypes Using Machine Learning Methods.

Simple SummaryMachine learning has been extensively used in analyzing big data and in conditions where the number of parameters is much bigger than the number of observations. Recently, there have been an increasing number of successful applications of machine learning in genomic prediction as this method makes weaker assumptions, is capable of dealing with the dimensionality problem, and can be more flexible for describing complex relationships. In this study, we evaluated the predictive ability of three machine learning methods, namely, random forest (RF), extreme gradient boosting (XGB), and support vector machine (SVM), when predicting the carcass traits of Hanwoo cattle. These machine learning algorithms were compared with the standard linear method (GBLUP). Our results revealed that XGB method had the best predictive correlation for carcass weight and marbling score. Meanwhile, the best predictive correlation for backfat thickness and eye muscle area was delivered by GBLUP. Moreover, in terms of mean squared error (MSE) of prediction, GBLUP delivered the lowest MSE value for all traits.Hanwoo was originally raised for draft purposes, but the increase in local demand for red meat turned that purpose into full-scale meat-type cattle rearing; it is now considered one of the most economically important species and a vital food source for Koreans. The application of genomic selection in Hanwoo breeding programs in recent years was expected to lead to higher genetic progress. However, better statistical methods that can improve the genomic prediction accuracy are required. Hence, this study aimed to compare the predictive performance of three machine learning methods, namely, random forest (RF), extreme gradient boosting method (XGB), and support vector machine (SVM), when predicting the carcass weight (CWT), marbling score (MS), backfat thickness (BFT) and eye muscle area (EMA). Phenotypic and genotypic data (53,866 SNPs) from 7324 commercial Hanwoo cattle that were slaughtered at the age of around 30 months were used. The results showed that the boosting method XGB showed the highest predictive correlation for CWT and MS, followed by GBLUP, SVM, and RF. Meanwhile, the best predictive correlation for BFT and EMA was delivered by GBLUP, followed by SVM, RF, and XGB. Although XGB presented the highest predictive correlations for some traits, we did not find an advantage of XGB or any machine learning methods over GBLUP according to the mean squared error of prediction. Thus, we still recommend the use of GBLUP in the prediction of genomic breeding values for carcass traits in Hanwoo cattle.

Accuracy of genomic evaluation using imputed high-density genotypes for carcass traits in commercial Hanwoo population

This study aimed to evaluate the accuracy of genomic prediction using single-trait and multi-trait single-step genomic best linear unbiased prediction (ssGBLUP) based on 50 K SNP chip dataset and imputed high-density (HD) genotypes for various carcass traits in commercial Hanwoo population such as backfat thickness (BFT), carcass weight (CWT), rib eye area (REA), and marbling score (MS). Analyses involved phenotypes from 18,269 animals born between April 2006 and June 2017 from different commercial herds in South Korea as well as genomic information from 16,892 animals genotyped using customized Hanwoo 50 K SNP chip and imputed to HD genotypes. Cross-validation was performed on 3,041 animals in the validation population born from November 2016 to June 2017. The results showed that HD genotypes led to a marginal increase in prediction accuracies (0.6 to 2%) than 50 K genotypes for BFT, REA, and MS, while no improvement was noted for CWT. Compared to the single-trait model, the use of multi-trait model based on 50 K or HD genotypes produced only a small improvement in prediction accuracies (1.2 to 2%) for BFT and REA but without improvement for CWT. For MS, the accuracy of genomic prediction using the multi-trait model based on 50 K or HD genotype datasets was slightly lower than the single-trait model. Generally, inflation of predictions increased using the HD genotypes and multi-trait model. Therefore, we recommend the use of 50 K genotypes and single-trait model in the estimation of genomic breeding values for carcass traits in commercial Hanwoo population.

Genetic evaluation of tropical climate-adapted sheep for carcass traits including genomic information

Abstract Santa Ines is a locally adapted hair sheep breed that is under large expansion in Brazil and other Tropical countries. This study aimed to evaluate the effect of different relationship matrices (A or H) on the estimation of genetic parameters and accuracy of (genomic) breeding values [(G)EBVs] for carcass traits measured in vivo in Santa Ines sheep. A total of 977, 890, 894, and 882 records for loin eye area (LEA), marbling score (MLE), subcutaneous fat thickness (SFT), and leg circumference (LEC), respectively, were included in this study. There were 1637 animals in the numerator relationship matrix (A). A total of 389 animals with phenotypic information were genotyped using the OvineSNP50 BeadChip. After the genotyping quality control, 388 samples and 42,748 SNPs remained in the dataset. The (co)variance components were estimated via Bayesian inference through single- and multi-trait analyses using the animal model with either the A or H matrix. The theoretical accuracies of (G)EBVs, Pearson correlation between breeding values estimated including or not genomic information, and Spearman rank correlation were used to evaluate the feasibility of incorporating genomic information in the analyses. Heritability estimates ranged from 0.12 ± 0.06 (SFT, using the A matrix) to 0.29 ± 0.07 (LEA, using the H matrix), in the single-trait analyses, and from 0.17 ± 0.05 (SFT, using A) to 0.33 ± 0.07 (LEA, using H), in a multi-trait setting. Direct genetic additive variance and heritability estimates were higher when including genomic information. Unfavorable and high genetic correlation was observed between LEA and SFT. The estimates of theoretical accuracy of (G)EBVs were higher in all the scenarios when genomic information was incorporated in the model. Gains in accuracy ranged from 0.013 to 0.039 units when genomic information was included in the analyses. Thus, if available, genomic information should be incorporated in the estimation of genetic parameters and breeding values for carcass traits measured in vivo in Santa Ines sheep.

Sensitivity of breeding values for carcass traits of meat-type quail to changes in dietary (methionine + cystine):lysine ratio using reaction norm models.

Our objective was to evaluate changes in breeding values for carcass traits of two meat-type quail (Coturnix coturnix) strains (LF1 and LF2) to changes in the dietary (methionine+cystine):lysine ([Met+Cys]:Lys) ratio due to genotype by environment (G×E) interaction via reaction norm. A total of 7000 records of carcass weight and yield were used for analyses. During the initial phase (from hatching to day 21), five diets with increasing (Met+Cys):Lys ratios (0.61, 0.66, 0.71, 0.76 and 0.81), containing 26.1% crude protein and 2900kcalME/kg, were evaluated. Analyses were performed using random regression models that included linear functions of sex (fixed effect) and breeding value (random effect) for carcass weight and yield, without and with heterogeneous residual variance adjustment. Both fixed and random effects were modelled using Legendre polynomials of second order. Genetic variance and heritability estimates were affected by both (Met+Cys):Lys ratio and strain. We observed that a G×E interaction was present, with changes in the breeding value ranking. Therefore, genetic evaluation for carcass traits should be performed under the same (Met+Cys):Lys ratio in which quails are raised.

Genomic prediction of breeding values for carcass traits in Nellore cattle.

BackgroundThe objective of this study was to evaluate the accuracy of genomic predictions for rib eye area (REA), backfat thickness (BFT), and hot carcass weight (HCW) in Nellore beef cattle from Brazilian commercial herds using different prediction models.MethodsPhenotypic data from 1756 Nellore steers from ten commercial herds in Brazil were used. Animals were offspring of 294 sires and 1546 dams, reared on pasture, feedlot finished, and slaughtered at approximately 2 years of age. All animals were genotyped using a 777k Illumina Bovine HD SNP chip. Accuracy of genomic predictions of breeding values was evaluated by using a 5-fold cross-validation scheme and considering three models: Bayesian ridge regression (BRR), Bayes C (BC) and Bayesian Lasso (BL), and two types of response variables: traditional estimated breeding value (EBV), and phenotype adjusted for fixed effects (Y*).ResultsThe prediction accuracies achieved with the BRR model were equal to 0.25 (BFT), 0.33 (HCW) and 0.36 (REA) when EBV was used as response variable, and 0.21 (BFT), 0.37 (HCW) and 0.46 (REA) when using Y*. Results obtained with the BC and BL models were similar. Accuracies increased for traits with a higher heritability, and using Y* instead of EBV as response variable resulted in higher accuracy when heritability was higher.ConclusionsOur results indicate that the accuracy of genomic prediction of carcass traits in Nellore cattle is moderate to high. Prediction of genomic breeding values from adjusted phenotypes Y* was more accurate than from EBV, especially for highly heritable traits. The three models considered (BRR, BC and BL) led to similar predictive abilities and, thus, either one could be used to implement genomic prediction for carcass traits in Nellore cattle.

Comparison of dimensionality reduction methods to predict genomic breeding values for carcass traits in pigs.

A significant contribution of molecular genetics is the direct use of DNA information to identify genetically superior individuals. With this approach, genome-wide selection (GWS) can be used for this purpose. GWS consists of analyzing a large number of single nucleotide polymorphism markers widely distributed in the genome; however, because the number of markers is much larger than the number of genotyped individuals, and such markers are highly correlated, special statistical methods are widely required. Among these methods, independent component regression, principal component regression, partial least squares, and partial principal components stand out. Thus, the aim of this study was to propose an application of the methods of dimensionality reduction to GWS of carcass traits in an F2 (Piau x commercial line) pig population. The results show similarities between the principal and the independent component methods and provided the most accurate genomic breeding estimates for most carcass traits in pigs.

Relationship between commercially available DNA analysis and phenotypic observations on beef quality and tenderness

Warner–Bratzler shear force values from 560 mixed breed heifers and steers were used to determine estimates of genetic selection. Cattle were marketed from 2008 to 2011, and included five feedlot based research projects at the North Dakota State University-Carrington Research Extension Center. Samples were collected for IGENITY® analysis providing information that included selection indices and estimated breeding values for carcass traits. DNA-based test results were compared with actual carcass measurements. Marbling accounted for over 10% of the variation in WBSF while hot carcass weight was the second most influential carcass trait accounting for 4% (P<0.01). Regression coefficients of IGENITY® molecular breeding value on phenotype for WBSF, marbling, ribeye area, yield grade, and fat thickness were low (R2=0.14, 0.02, 0.03, 0.03, and 0.02, respectively). Therefore selecting cattle for a higher degree of marbling and feeding a diet that meets or exceeds recommended nutrients for growth are the most important factors influencing beef tenderness and acceptability.

Breeding Values for Carcass Traits at Calf Markets as Determinant of Feeding Length in Japanese Black Cattle: an Exploratory Study

Japanese Black cattle (Wagyu) are fed for a long period to produce high quality beef, however, extended feeding often causes inefficiency and greater environmental load mainly derived from their manure. The objectives of this study were to analyze changes in feeding length by listing breeding values (BVs) at calf markets and the relationships between BVs and carcass characteristics of 4,052 Japanese Black cattle, and to examine the feasibility of optimizing feeding length by referring to listed BVs. BV classes A, B, and C were defined based on BVs of cows in Shimane Prefecture as follows: an upper quarter of BVs was classified as A, a second quarter as B, and under the average as C. For cattle sold at calf markets in the first term of 1996, just before the start of BV listing, the feeding length of cattle with class B BVs for the beef marbling standard (BMS) was longer (p>0.05) than that of class A cattle. However, in the second term of 1996, just after the start of BV listing, the feeding length of class B cattle became shorter (p<0.001) than that of class A cattle. Then, the feeding lengths of both classes showed no significant differences. Feeding lengths of both class A and B BVs for carcass weight (CW) changed similarly to the corresponding BV classes for BMS. The analysis of the relationships among the listed BV classes and the actual carcass characteristics showed that class A cattle had a higher (p<0.001) BMS than class B cattle, and that the higher-class cattle had a heavier CW (p<0.05). On the basis of previous reports, the cattle, particularly those with lower genetic marbling ability, seem to only increase marbling at markedly low efficiency for a few months before slaughter. Therefore, the finding that carcass characteristics corresponded to their class of BVs suggests that an optimum feeding length based on listed BVs not only increases the efficiency of beef production, but also reduces the environmental load. (Asian-Aust. J. Anim. Sci. 2003. Vol 16, No. 5 : 635-638)