Design Of A Genome-wide Low-density Panel On Important Growth Traits In Simmental Cattle

As a very current topic in livestock production, genomic-enabled prediction has been widely used in animal breeding and it improves the production dramatically. However, the high cost and huge burden of calculation restrict its development. Therefore, as an alternative, low density panels with SNPs that show strong associations with phenotype could be used.The reference population consist of 1,059 Simmental cattle which were born in 2008 and 2012. All these samples had been genotyped by Illumina 770 K SNP and three economical traits were selected to testify the efficiency of the genomic selection with low density markers, which are body weight, carcass weight and average daily gain. Different markers densities were constructed based on the evenly-spaced, the ranking of BayesB result and selected SNPs from genome wide association study(GWAS) result, respectively. Consistent with that, three low density datasets(ELD, SLD and PLD) were left for the subsequently cross validation work to evaluate the accuracy of genomic selection with low density markers.It can be seen from the results that with the increase of the markers’ number, the accuracy of three low density datasets all show a consistently increasing trend. Among them, the strategy with the BayesB obtained the higher accuracy, compared with the other two methods. Especially, when the number of SNP attains 10,000, the accuracy of genomic prediction with SLD reaches 0.22±0.01 for body weight, 0.21±0.02 for carcass weight and 0.15±0.01 for average daily gain, correpondingly. In terms of three different traits, the predictive ability performs slightly different. In the prediction of body weight and the average daily gain, despite the relatively lower predictive ability with GBLUP method, the left were all beyond the other two low density chips. It means that the predictive ability of low density markers may be related with the genetic architecture and its application still needs detailed analysis of breeding aim. Using the Simmental cattle as the reference population and sepecialized with body weight, carcass weight, as well as average daily gain, we proposed three different types of low density chips and evaluated their accuracy. Our study gives an insight into the feasiable and efficient genomic selection with low density markers in cattle.