Short communication: The combined use of linkage disequilibrium-based haploblocks and allele frequency-based haplotype selection methods enhances genomic evaluation accuracy in dairy cattle

摘要

The construction and use of haploblocks [adjacent single nucleotide polymorphisms (SNP) in strong link-age disequilibrium] for genomic evaluation is advanta-geous, because the number of effects to be estimated can be reduced without discarding relevant genomic information. Furthermore, haplotypes (the combina-tion of 2 or more SNP) can increase the probability of capturing the quantitative trait loci effect compared with individual SNP markers. With regards to haplo-types, the allele frequency parameter is also of interest, because as a selection criterion, it allows the number of rare alleles to be reduced, and the effects of those alleles are usually difficult to estimate. We have proposed a simple pipeline that simultaneously incorporates link-age disequilibrium and allele frequency information in genomic evaluation, and here we present the first results obtained with this procedure. We used a population of 2,235 progeny-tested bulls from the Montbeliarde breed for the tests. Phenotype data were available in the form of daughter yield deviations on 5 production traits, and genotype data were available from the 50K SNP chip. We conducted a classical validation study by splitting the population into training (80% oldest animals) and validation (20% youngest animals) sets to emulate a real-life scenario in which the selection candidates had no available phenotype data. We measured all reported parameters for the validation set. Our results proved that the proposed method was indeed advantageous, and that the accuracy of genomic evaluation could be improved. Compared with results from a genomic BLUP analysis, correlations between daughter yield deviations (a proxy for true) and genomic estimated breeding val-ues increased by an average of 2.7 percentage points for the 5 traits. Inflation of the genomic evaluation of the selection candidates was also significantly reduced. The proposed method outperformed the other SNP and haplotype-based tests we had evaluated in a previous study. The combination of linkage disequilibrium-based haploblocks and allele frequency-based haplotype se-lection methods is a promising way to improve the ef-ficiency of genomic evaluation. Further work is needed to optimize each step in the proposed analysis pipeline.