High production ability has been used for primary selection in dairy breeding schemes. In particular, milk yield, fat yield, and protein yield are the most important economic traits for dairy cattle selection. To date, genetic improvement of these economic traits has been performed successfully based on traditional best linear unbiased prediction (BLUP), and the breeding values of economic traits have been applied with selection indices in Korean dairy breeding systems. The BLUP used in combination with individual records and estimated breeding value (EBV) has resulted in considerable genetic progress in the dairy industry. In recent years, however, genomic information in the form of commercial single-nucleotide polymorphic (SNP) marker panels from various companies (i.e., Illumina, San Diego, CA, USA; Neogen-GeneSeek, Lincoln, NE, USA; and Affymetrix, Santa Clara, CA, USA) have become available for genetic evaluations, as a consequence of improvements in genotyping technology and statistical methods after introduction by Meuwissen et al in 2001. Accordingly, genomic prediction using genotypic data has been widely applied for various livestock.
Genomic selection (GS) involves selection of bulls based on genomic breeding values, which are derived from the combination of EBVs and direct breeding values (DGVs) based on SNPs using several blending formulae or single-step methods (e.g., ss-GBLUP (single-step genomic Best Linear Unbiased Prediction) and single-step super hybrid model.
The advantages of GS are simplicity and resistance to preselection bias and more reliable prediction than traditional BLUP. When GS schemes are applied in the field, the use of young bulls should be the most effective in terms of reliability. For example, in young Holstein bulls in the United States, reliabilities for predicted transmitting abilities for milk yield based on genomic information ranged from 73% to 79%.
The most important thing in GS is reference population that has both phenotype and genomic information. The larger the reference population size, the higher the accuracy of the GS. However, the total dairy population in Korea is small with about 400,000 heads, so there is a limit to growing the reference population. In Korea, it is difficult to expect the effects of GS as much as countries with large reference size. The other important thing in GS is using genetic information that including how many significant makers. Previous studies have shown that using custom array with significant markers for dairy cow is more effective than using conventional commercial array (e.g., Illumina bovine 50k array).
The objective of using custom array (Axiom myDesign Array plate) were to increase genomic accuracy while overcoming the small population size by including significant markers for Korean dairy cattle in the custom array.