Search

Background Genomic prediction faces two main statistical problems: multicollinearity and n¿«¿p (many fewer observations than predictor variables). Principal component (PC) analysis is a multivariate statistical method that is often used to address these problems. The objective of this study was to compare the performance of PC regression (PCR) for genomic prediction with that of a commonly used REML model with a genomic relationship matrix (GREML) and to investigate the full potential of PCR for genomic prediction. Methods The PCR model used either a common or a semi-supervised approach, where PC were selected based either on their eigenvalues (i.e. proportion of variance explained by SNP (single nucleotide polymorphism) genotypes) or on their association with phenotypic variance in the reference population (i.e. the regression sum of squares contribution). Cross-validation within the reference population was used to select the optimum PCR model that minimizes mean squared error. Pre-corrected average daily milk, fat and protein yields of 1609 first lactation Holstein heifers, from Ireland, UK, the Netherlands and Sweden, which were genotyped with 50 k SNPs, were analysed. Each testing subset included animals from only one country, or from only one selection line for the UK. Results In general, accuracies of GREML and PCR were similar but GREML slightly outperformed PCR. Inclusion of genotyping information of validation animals into model training (semi-supervised PCR), did not result in more accurate genomic predictions. The highest achievable PCR accuracies were obtained across a wide range of numbers of PC fitted in the regression (from one to more than 1000), across test populations and traits. Using cross-validation within the reference population to derive the number of PC, yielded substantially lower accuracies than the highest achievable accuracies obtained across all possible numbers of PC. Conclusions On average, PCR performed only slightly less well than

For several decades, breeding goals in dairy cattle focussed on increased milk production. However, many functional traits have negative genetic correlations with milk yield, and reductions in genetic merit for health and fitness have been observed. Herd management has been challenged to compensate for these effects and to balance fertility, udder health and metabolic diseases against increased production to maximize profit without compromising welfare. Functional traits, such as direct information on cow health, have also become more important because of growing concern about animal well-being and consumer demands for healthy and natural products. There are major concerns about the impact of drugs used in veterinary medicine on the spread of antibiotic-resistant strains of bacteria that can negatively impact human health. Sustainability and efficiency are also increasingly important because of the growing competition for high-quality, plant-based sources of energy and protein. Disruptions to global environments because of climate change may encourage yet more emphasis on these traits. To be successful, it is vital that there be a balance between the effort required for data recording and subsequent benefits. The motivation of farmers and other stakeholders involved in documentation and recording is essential to ensure good data quality. To keep labour costs reasonable, existing data sources should be used as much as possible. Examples include the use of milk composition data to provide additional information about the metabolic status or energy balance of the animals. Recent advances in the use of mid-infrared spectroscopy to measure milk have shown considerable promise, and may provide cost-effective alternative phenotypes for difficult or expensive-to-measure traits, such as feed efficiency. There are other valuable data sources in countries that have compulsory documentation of veterinary treatments and drug use. Additional sources of data outside of the farm in

The main objective was to evaluate the potential of grass silages of very high quality to support a high milk yield with a low or moderate, or even without concentrate supplementation. Production responses to increased levels of concentrate supplementation with 3 primary growth grass silages differing in digestibility were studied using 66 Norwegian Red dairy cows. Roundbale silage was produced from a timothy-dominated sward at very early (H1), early (H2), and normal (H3) stages of crop maturity. Crops were rapidly wilted (<24h) and a formic acid-based additive was applied. All silages were restrictedly fermented. Silage digestible organic matter in dry matter (DM) values were 747, 708, and 647 g/kg of DM for H1, H2, and H3, respectively. Dietary treatments were fed in a 3×3 factorial arrangement of the 3 silages supplemented with 3 concentrate levels (4, 8, and 12 kg/d) and, additionally, H1 was offered without concentrates and H3 with 16 kg/d, giving a total of 11 diets. Cows, blocked according to parity and calving date, were introduced to the experiment before calving and kept in the experiment until wk 16 of lactation. Silage was offered ad libitum in loose housing and concentrate was available in automatic feed stations. Intake of grass silage when fed as the sole feed was 16.9 kg of DM on average for lactation wk 1 to 16. When H1 was supplemented with 4 or 8 kg of concentrates, silage DM intake did not change, but total DM intake increased to 20.6 and 23.7 kg/d, respectively. Energy-corrected milk (ECM) yield increased from 23.4 kg when H1 was offered without concentrate supplement to 29.1 and 32.8 kg when supplemented with 4 or 8 kg concentrate, respectively. None of the other diets equaled the yield obtained by H1 plus 8 kg of concentrate. Feed intake and yield of cows offered H3 plus 4 kg of concentrates were strongly constrained by high dietary fiber concentration. They consumed 16.5 g of neutral detergent fiber/kg of body weight and spent more time ea