ACTA VETERINARIA ET ZOOTECHNICA SINICA ›› 2017, Vol. 48 ›› Issue (1): 60-67.doi: 10.11843/j.issn.0366-6964.2017.01.007

Previous Articles     Next Articles

Bayesian Models including Dominant Effects for Genomic Selection

WANG Yan-hui, ZHU Bo, LI Jun-ya*   

  1. (Institute of Animal Science, Chinese Academy of Agricultural Sciences, Beijing 100193,China)
  • Received:2016-10-27 Online:2017-01-23 Published:2017-01-23

Abstract:

The aim of this study was to investigate the impact of dominant effects on the predictive accuracy of genomic breeding values. Based on the dependency between additive and dominant effects using BayesA model, we proposed two submodels:1) the additive effect and dominant effect were independent of each other in BayesAD1 model; 2) the dominant coefficient and absolute values of additive effect were independent of each other in BayesAD2 model, in which the dominant coefficient followed normal distribution. Using simulated datasets, we compared the predictive accuracy of genomic estimated breeding value (GEBV) among the additive model (BayesAD0) and dominant models (BayesAD1 and BayesAD2). We further investigated the effect of number of QTLs (quantitative trait loci), size of full-sibs family and ratio of additive variance to dominant variance on predictive accuracy of GEBV. The results showed that the dominant models slowed down the declining of the predictive accuracy of GEBV in subsequent generations. Moreover, the predictive accuracy of GEBV increased as the ratio of dominant variance increase. When the ratio of additive variance to dominant variance reached 0.25, BayesAD2 was 20.3% and 28.4% higher than the accuracy of the BayesAD1 and BayesAD0, respectively. In addition, the size of full-sibs family affected the predictive accuracy of GEBV positively. And an increase in the number of QTL was accompanied by a reduction on the predictive accuracy of GEBV. These results indicate that a better prediction of genetic values is intended, when the dominant variance are large just as low-heritability traits.

CLC Number: