The superiority of multi-trait models with genotype-by-environment interactions in a limited number of environments for genomic prediction in pigs

Song, Huihui; Zhang, Qin; Ding, Xiangdong

doi:10.1186/s40104-020-00493-8

Cited by 15 publications

(15 citation statements)

References 35 publications

(43 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Similar results have been reported for growth and reproduction traits in Yorkshire pigs (Song et al, 2017) and fatty acid composition traits in Korean Hanwoo cattle (Bhuiyan et al, 2018). Moreover, as reported from studies on real data (Hayes et al, 2009;Song et al, 2020;Zenger et al, 2019), the average prediction accuracies were not significantly different between the GBLUP and Bayesian methods, and the Bayesian method was computationally very demanding. The performance of Bayesian F I G U R E 1 Accuracy of genomic prediction using GBLUP with different densities of SNP panels in four populations methods depends on their assumptions and the underlying genetic architecture of the traits.…”

Section: Discussionsupporting

confidence: 85%

Strategies to improve the accuracy and reduce costs of genomic prediction in aquaculture species

Song¹,

Hu²

2021

Evolutionary Applications

Self Cite

View full text Add to dashboard Cite

Genomic selection (GS) has great potential to increase genetic gain in aquaculture breeding; however, its implementation is hindered owing to high genotyping cost and the large number of individuals to genotype. This study investigated the efficiency of genomic prediction in four aquaculture species. In total, 749 to 1481 individuals with records for disease resistance and growth traits were genotyped using SNP arrays ranging from 12K to 40K. We compared the prediction accuracies and bias of breeding values obtained from BLUP, genomic BLUP (GBLUP), Bayesian mixture (BayesR), weighted GBLUP (WGBLUP), and genomic feature BLUP (GFBLUP). For GFBLUP, the genomic feature matrix was constructed based on prior information from genome‐wide association studies. Fivefold cross‐validation was performed with 20 replicates. Moreover, to reduce the cost of GS, we reduced the SNP density based on linkage disequilibrium as well as the reference population size. The results showed that the methods with marker information produced more accurate predictions than the pedigree‐based BLUP method. For the genomic model, BayesR performed prediction with a similar or higher accuracy compared to GBLUP. For the four traits, WGBLUP yielded an average of 1.5% higher accuracy than GBLUP. However, the accuracy of genomic prediction decreased by an average of 6.2% for GFBLUP compared to GBLUP. When the density of SNP panels was reduced to 3K, which was sufficient to obtain accuracies similar to those using the whole dataset in the four species, the cost of GS was estimated to be 50% lower than that of genotyping all animals with high‐density panels. In addition, when the reference population size was reduced by 10%, evenly from full‐sib family, the accuracy of genomic prediction was almost unchanged, and the cost reduction was 8% in the four populations. Our results have important implications for translating the benefits of GS to most aquaculture species.

show abstract

Section: Discussionsupporting

confidence: 85%

Strategies to improve the accuracy and reduce costs of genomic prediction in aquaculture species

Song¹,

Hu²

2021

Evolutionary Applications

Self Cite

View full text Add to dashboard Cite

show abstract

“…(II) The low heritability of TNB and NBA. In this study, the heritability for the two traits were both 0.12, which was generally consistent with other reports [ 27 , 41 , 42 ]; therefore, sufficient accuracy could not be achieved with the pedigree information. This also confirmed by other studies, that a certain improvement can be achieved by adding a smaller reference population for traits with medium or high heritability [ 2 , 43 ].…”

Section: Discussionsupporting

confidence: 90%

Using machine learning to improve the accuracy of genomic prediction of reproduction traits in pigs

Wang

Shi

Wang³

et al. 2022

J Animal Sci Biotechnol

Self Cite

View full text Add to dashboard Cite

Background Recently, machine learning (ML) has become attractive in genomic prediction, but its superiority in genomic prediction over conventional (ss) GBLUP methods and the choice of optimal ML methods need to be investigated. Results In this study, 2566 Chinese Yorkshire pigs with reproduction trait records were genotyped with the GenoBaits Porcine SNP 50 K and PorcineSNP50 panels. Four ML methods, including support vector regression (SVR), kernel ridge regression (KRR), random forest (RF) and Adaboost.R2 were implemented. Through 20 replicates of fivefold cross-validation (CV) and one prediction for younger individuals, the utility of ML methods in genomic prediction was explored. In CV, compared with genomic BLUP (GBLUP), single-step GBLUP (ssGBLUP) and the Bayesian method BayesHE, ML methods significantly outperformed these conventional methods. ML methods improved the genomic prediction accuracy of GBLUP, ssGBLUP, and BayesHE by 19.3%, 15.0% and 20.8%, respectively. In addition, ML methods yielded smaller mean squared error (MSE) and mean absolute error (MAE) in all scenarios. ssGBLUP yielded an improvement of 3.8% on average in accuracy compared to that of GBLUP, and the accuracy of BayesHE was close to that of GBLUP. In genomic prediction of younger individuals, RF and Adaboost.R2_KRR performed better than GBLUP and BayesHE, while ssGBLUP performed comparably with RF, and ssGBLUP yielded slightly higher accuracy and lower MSE than Adaboost.R2_KRR in the prediction of total number of piglets born, while for number of piglets born alive, Adaboost.R2_KRR performed significantly better than ssGBLUP. Among ML methods, Adaboost.R2_KRR consistently performed well in our study. Our findings also demonstrated that optimal hyperparameters are useful for ML methods. After tuning hyperparameters in CV and in predicting genomic outcomes of younger individuals, the average improvement was 14.3% and 21.8% over those using default hyperparameters, respectively. Conclusion Our findings demonstrated that ML methods had better overall prediction performance than conventional genomic selection methods, and could be new options for genomic prediction. Among ML methods, Adaboost.R2_KRR consistently performed well in our study, and tuning hyperparameters is necessary for ML methods. The optimal hyperparameters depend on the character of traits, datasets etc.

show abstract

“…(II) The low heritabilities of TNB and NBA. In this study, the heritabilities for the two traits were both 0.12, which was generally consistent with other reports [25,42,43] , therefore, it cannot get enough accuracy from the pedigree information.…”

Section: Discussionsupporting

confidence: 90%

Using Machine Learning To Improve the Accuracy of Genomic Prediction on Reproduction Traits in Pigs

Wang

Shi

Wang³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Background Recently, machine learning (ML) is becoming attractive in genomic prediction, while its superiority in genomic prediction and the choosing of optimal ML methods are needed investigation. Results In this study, 2566 Chinese Yorkshire pigs with reproduction traits records were used, they were genotyped with GenoBaits Porcine SNP 50K and PorcineSNP50 panel. Four ML methods, including support vector regression (SVR), kernel ridge regression (KRR), random forest (RF) and Adaboost.R2 were implemented. Through 20 replicates of five-fold cross-validation, the genomic prediction abilities of ML methods were explored. Compared with genomic BLUP(GBLUP), single-step GBLUP (ssGBLUP) and Bayesian method BayesHE, our results indicated that ML methods significantly outperformed. The prediction accuracy of ML methods was improved by 19.3%, 15.0% and 20.8% on average over GBLUP, ssGBLUP and BayesHE, ranging from 8.9–24.0%, 7.6–17.5% and 11.1–24.6%, respectively. In addition, ML methods yielded smaller mean squared error (MSE) and mean absolute error (MAE) in all scenarios. ssGBLUP yielded improvement of 3.7% on average compared to GBLUP, and the performance of BayesHE was close to GBLUP. Among four ML methods, SVR and KRR had the most robust prediction abilities, which yielded higher accuracies, lower bias, lower MSE and MAE, and comparable computing efficiency as GBLUP. RF demonstrated the lowest prediction ability and computational efficiency among ML methods. Conclusion Our findings demonstrated that ML methods are more efficient than traditional genomic selection methods, and it could be new options for genomic prediction.

show abstract

The superiority of multi-trait models with genotype-by-environment interactions in a limited number of environments for genomic prediction in pigs

Cited by 15 publications

References 35 publications

Strategies to improve the accuracy and reduce costs of genomic prediction in aquaculture species

Strategies to improve the accuracy and reduce costs of genomic prediction in aquaculture species

Using machine learning to improve the accuracy of genomic prediction of reproduction traits in pigs

Using Machine Learning To Improve the Accuracy of Genomic Prediction on Reproduction Traits in Pigs

Contact Info

Product

Resources

About