Advantage of including Genomic Information to Predict Breeding Values for Lactation Yields of Milk, Fat, and Protein or Somatic Cell Score in a New Zealand Dairy Goat Herd

Abstract

Selection on genomic breeding values (GBVs) is now readily available for ranking candidates in improvement schemes. Our objective was to quantify benefits in terms of accuracy of prediction from including genomic information in the single-trait estimation of breeding values (BVs) for a New Zealand mixed breed dairy goat herd. The dataset comprised phenotypic and pedigree records of 839 does. The phenotypes comprised estimates of 305-day lactation yields of milk, fat, and protein and average somatic cell score from the 2016 production season. Only 388 of the goats were genotyped with a Caprine 50K SNP chip and 41,981 of the single nucleotide polymorphisms (SNPs) passed quality control. Pedigree-based best linear unbiased prediction (PBLUP) was used to obtain across-breed breeding values (EBVs), whereas a single-step BayesC model (ssBC) was used to estimate across-breed GBVs. The average prediction accuracies ranged from 0.20 to 0.22 for EBVs and 0.34 to 0.43 for GBVs. Accuracies of GBVs were up to 103% greater than EBVs. Breed effects were more reliably estimated in the ssBC model compared with the PBLUP model. The greatest benefit of genomic prediction was for individuals with no pedigree or phenotypic records. Including genomic information improved the prediction accuracy of BVs compared with the current pedigree-based BLUP method currently implemented in the New Zealand dairy goat population.

Changes in the milk response to different herbage mass of dairy cattle in a tropical climate

“…Milk protein concentration is a complex variable to modify (Walker et al, 2004), which occurred in this study, where changes in the HM could not cause a high difference in ME intake, which has been positively related to the milk protein concentration (Walker et al, 2004). Although ME intake was greater for LHM, it was not enough to modify the milk protein, given that it is a variable that, besides being measured by the genetic merit (Scholtens et al, 2020), needs large amounts of non-structural carbohydrates to increase its concentration in milk (Mordenti et al, 2021). Nevertheless, a slight raise in milk protein concentration can positively affect economic profitability, given that it is the most expensive nutrient paid by the dairy industry.…”

Progress and Application of Genomic Selection in Sheep and Goat

“…Lillehammer (Lillehammer et al, 2020) conducted GS simulation studies under different models for carcass traits, growth traits and maternal traits of Norwegian white sheep and found that GS increased the genetic benefits of these traits by 18% to 20%. Scholtens (Scholtens et al, 2020) used Bayes C to estimate the GEBV of milk yield, protein content, and somatic score of New Zealand dairy goat, and the average prediction accuracy was 0.34 to 0.43, which was more reliable than the traditional BLUP. Ashraf (Ashraf et al, 2022) used GBLUP and Bayes R, Bayes A, Bayes B, Bayes L to perform GS and compared eight different traits of Soay Sheep, such as body weight, length of front and rear legs, coat color, patterns and so on.…”

Validation with single-step SNPBLUP shows that evaluations can continue using a single mean of genotyped individuals, even with multiple breeds

“…We found a single published example of a separate J-factor being fitted per breed but no indication of whether there was any benefit [ 12 ]. Our first aim for this study was to determine if fitting a J-factor for each breed in a crossbred population would have any benefit on accuracy of prediction or its dispersion bias, with a particular focus on when selective genotyping is applied.…”