Genetics Research

Short Note

Accuracy of genomic selection using stochastic search variable selection in Australian Holstein Friesian dairy cattle

KLARA L. VERBYLAa1a2a3 c1, BEN J. HAYESa1, PHILIP J. BOWMANa1 and MICHAEL E. GODDARDa1a2a3

a1 Biosciences Research Division, Department of Primary Industries Victoria, 1 Park Drive, Bundoora 3083, Australia

a2 Melbourne School of Land and Environment, The University of Melbourne, Parkville 3010, Australia

a3 The Cooperative Research Centre for Beef Genetic Technologies, University of New England, Armidale, NSW 2351, Australia

Summary

Genomic selection describes a selection strategy based on genomic breeding values predicted from dense single nucleotide polymorphism (SNP) data. Multiple methods have been proposed but the critical issue is how to decide whether an SNP should be included in the predictive set to estimate breeding values. One major disadvantage of the traditional Bayes B approach is its high computational demands caused by the changing dimensionality of the models. The use of stochastic search variable selection (SSVS) retains the same assumptions about the distribution of SNP effects as Bayes B, while maintaining constant dimensionality. When Bayesian SSVS was used to predict genomic breeding values for real dairy data over a range of traits it produced accuracies higher or equivalent to other genomic selection methods with significantly decreased computational and time demands than Bayes B.

(Received July 21 2009)

(Revised September 10 2009)

Correspondence:

c1 Corresponding author. e-mail: klara.verbyla@dpi.vic.gov.au

Metrics