|Hay, El Hamidi|
|UTSUNOMIYA, YURI - Universidade Estadual Paulista (UNESP)|
|XU, LINGYANG - Collaborator|
|ZHOU, YANG - Northwest Agriculture And Forestry University|
|NEVES, HAROLDO - Universidade Estadual Paulista (UNESP)|
|CARVALHEIRO, ROBERTO - Universidade Estadual Paulista (UNESP)|
|GARCIA, JOSE FERNANDO - Universidade Estadual Paulista (UNESP)|
|Liu, Ge - George|
Submitted to: Biomed Central (BMC) Genomics
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/14/2018
Publication Date: 6/5/2018
Citation: Hay, E.A., Utsunomiya, Y.T., Xu, L., Zhou, Y., Neves, H.H., Carvalheiro, R., Bickhart, D.M., Garcia, J., Liu, G. 2018. Genomic predictions combining SNP markers and copy number variations in Nellore cattle. Biomed Central (BMC) Genomics. 19(1):441. https://doi.org/10.1186/s12864-018-4787-6.
Interpretive Summary: Unlike other well-known genetic variations such as single nucleotide polymorphisms (SNPs), structural variations in the genome known as gene copy number variations (CNVs) have never been used in genomic prediction. In this study, we performed the first genomic predictions for cattle that incorporated CNV markers along with SNP markers. This approach can increase the accuracy of genomic estimated breeding values for some traits and lead to additional genetic gain during selection. Farmers, scientist, and policy planners working to improve animal health and production based on genome-enable animal selection will benefit from this study.
Technical Abstract: Background: Genomic selection is routinely used due to the advancement in high throughput technology to efficiently genotype large number of SNPs. However, genomic selection fails in the case of complex traits. SNP markers are not sufficient to predict these traits, therefore additional factors like copy number variations (CNVs) were considered. Copy number variations have been shown to affect several phenotypic traits. Results: In this study, CNVs were included in a genomic selection framework and modeled through a Bayesian hierarchical model. A Nellore cattle population consisting of 2,230 animals genotyped for 777K SNPs was used and 9 weight and carcass traits were analyzed. Three models were adopted to incorporate CNVs: 1) Bayesian model including SNP markers and CNVs similar to BayesA (BayesA_CNV); 2) Bayesian mixture model (BayesB_CNV); 3) genomic BLUP without polygenic effects (GBLUP_CNV). Accuracies were assessed based on the correlation between EBV and DGV in the validation dataset. For the first model and second model, accuracy ranged from 0.29 to 0.62. The accuracy significantly increased for some traits (CW, MW, PW, PWG, WG) when including CNVs in the model very likely due to better capturing of the genetic variation. The second model performed better resulting in an increase in accuracy ranging from 5 to 25%. Using GBLUP_CNV which considers CNVs as covariates have also resulted in an increase in prediction accuracy. Conclusions: This study presents the first CNV-based genomic prediction. Combining CNV and SNP marker information proved to be beneficial for several traits in Nellore cattle.