A.P.W. de Roos

Linkage Disequilibrium and Persistence of Phase in Holstein–Friesian, Jersey and Angus Cattle

When a genetic marker and a quantitative trait locus (QTL) are in linkage disequilibrium (LD) in one population, they may not be in LD in another population or their LD phase may be reversed. The objectives of this study were to compare the extent of LD and the persistence of LD phase across multiple cattle populations. LD measures r and r2 were calculated for syntenic marker pairs using genomewide single-nucleotide polymorphisms (SNP) that were genotyped in Dutch and Australian Holstein–Friesian (HF) bulls, Australian Angus cattle, and New Zealand Friesian and Jersey cows. Average r2 was ∼0.35, 0.25, 0.22, 0.14, and 0.06 at marker distances 10, 20, 40, 100, and 1000 kb, respectively, which indicates that genomic selection within cattle breeds with r2 ≥ 0.20 between adjacent markers would require ∼50,000 SNPs. The correlation of r values between populations for the same marker pairs was close to 1 for pairs of very close markers (<10 kb) and decreased with increasing marker distance and the extent of divergence between the populations. To find markers that are in LD with QTL across diverged breeds, such as HF, Jersey, and Angus, would require ∼300,000 markers.

Accuracy of Genomic Selection Using Different Methods to Define Haplotypes

Genomic selection uses total breeding values for juvenile animals, predicted from a large number of estimated marker haplotype effects across the whole genome. In this study the accuracy of predicting breeding values is compared for four different models including a large number of markers, at different marker densities for traits with heritabilities of 50 and 10%. The models estimated the effect of (1) each single-marker allele [single-nucleotide polymorphism (SNP)1], (2) haplotypes constructed from two adjacent marker alleles (SNP2), and (3) haplotypes constructed from 2 or 10 markers, including the covariance between haplotypes by combining linkage disequilibrium and linkage analysis (HAP_IBD2 and HAP_IBD10). Between 119 and 2343 polymorphic SNPs were simulated on a 3-M genome. For the trait with a heritability of 10%, the differences between models were small and none of them yielded the highest accuracies across all marker densities. For the trait with a heritability of 50%, the HAP_IBD10 model yielded the highest accuracies of estimated total breeding values for juvenile and phenotyped animals at all marker densities. It was concluded that genomic selection is considerably more accurate than traditional selection, especially for a low-heritability trait.

Reliability of Genomic Predictions Across Multiple Populations

Genomic prediction of future phenotypes or genetic merit using dense SNP genotypes can be used for prediction of disease risk, forensics, and genomic selection of livestock and domesticated plant species. The reliability of genomic predictions is their squared correlation with the true genetic merit and indicates the proportion of the genetic variance that is explained. As reliability relies heavily on the number of phenotypes, combining data sets from multiple populations may be attractive as a way to increase reliabilities, particularly when phenotypes are scarce. However, this strategy may also decrease reliabilities if the marker effects are very different between the populations. The effect of combining multiple populations on the reliability of genomic predictions was assessed for two simulated cattle populations, A and B, that had diverged for T = 6, 30, or 300 generations. The training set comprised phenotypes of 1000 individuals from population A and 0, 300, 600, or 1000 individuals from population B, while marker density and trait heritability were varied. Adding individuals from population B to the training set increased the reliability in population A by up to 0.12 when the marker density was high and T = 6, whereas it decreased the reliability in population A by up to 0.07 when the marker density was low and T = 300. Without individuals from population B in the training set, the reliability in population B was up to 0.77 lower than in population A, especially for large T. Adding individuals from population B to the training set increased the reliability in population B to close to the same level as in population A when the marker density was sufficiently high for the marker–QTL linkage disequilibrium to persist across populations. Our results suggest that the most accurate genomic predictions are achieved when phenotypes from all populations are combined in one training set, while for more diverged populations a higher marker density is required.

Production loss due to new subclinical mastitis in Dutch dairy cows estimated with a test-day model

Milk, fat, and protein loss due to a new subclinical mastitis case may be economically important, and the objective of this study was to estimate this loss. The loss was estimated based on test-day (TD) cow records collected over a 1-yr period from 400 randomly selected Dutch dairy herds. After exclusion of records from cows with clinical mastitis, the data set comprised 251,647 TD records from 43,462 lactations of 39,512 cows. The analysis was carried out using a random regression test-day modeling approach that predicts the cow production at each TD based on the actual production at all previous TD. The definition of new subclinical mastitis was based on the literature and assumed a new subclinical case if somatic cell count (SCC) was >100,000 cells/mL after a TD with SCC <50,000 cells/mL. A second data set was created by applying an adjustment to correct low SCC for the dilution effect when determining if the previous test-day SCC was <50,000 cells/ mL. Thereafter, the loss was estimated for records with SCC >100,000 cells/mL. The production (milk, fat, or protein) losses were modeled as the difference between the actual and predicted production (milk, fat, or protein) at the TD of new subclinical mastitis, for 4,382 cow records, and 2,545 cow records after dilution correction. Primiparous cows were predicted to lose 0.31 (0.25-0.37) and 0.28 (0.20-0.35) kg of milk/d at an SCC of 200,000 cells/mL, for unadjusted and adjusted low SCC, respectively. For the same SCC increase, multiparous cows were predicted to lose 0.58 (0.54-0.62) and 0.50 (0.44-0.56) kg of milk/d, respectively. Moreover, it was found that the greater the SCC increase above 100,000 cells/mL, the greater the production losses. The estimated production losses were more precise than previously reported estimates.

Imputation of genotypes from different single nucleotide polymorphism panels in dairy cattle

Imputation of missing genotypes is important to join data from animals genotyped on different single nucleotide polymorphism (SNP) panels. Because of the evolution of available technologies, economical reasons, or coexistence of several products from competing organizations, animals might be genotyped for different SNP chips. Combined analysis of all the data increases accuracy of genomic selection or fine-mapping precision. In the present study, real data from 4,738 Dutch Holstein animals genotyped with custom-made 60K Illumina panels (Illumina, San Diego, CA) were used to mimic imputation of genotypes between 2 SNP panels of approximately 27,500 markers each and with 9,265 SNP markers in common. Imputation efficiency increased with number of reference animals (genotyped for both chips), when animals genotyped on a single chip were included in the training data, with regional higher marker densities, with greater distance to chromosome ends, and with a closer relationship between imputed and reference animals. With 0 to 2,000 animals genotyped for both chips, the mean imputation error rate ranged from 2.774 to 0.415% and accuracy ranged from 0.81 to 0.96. Then, imputation was applied in the Dutch Holstein population to predict alleles from markers of the Illumina Bovine SNP50 chip with markers from a custom-made 60K Illumina panel. A cross-validation study performed on 102 bulls indicated that the mean error rate per bull was approximately equal to 1.0%. This study showed the feasibility to impute markers in dairy cattle with the current marker panels and with error rates below 1%.

Short communication: Ketone body concentration in milk determined by Fourier transform infrared spectroscopy: Value for the detection of hyperketonemia in dairy cows

The objective of this study was to evaluate Fourier transform infrared (FTIR) spectrometry to measure milk ketone bodies to detect hyperketonemic cows and compare this method with milk fat to protein ratio to detect hyperketonemia. Plasma and milk samples were obtained weekly from calving to wk 9 postpartum from 69 high-producing dairy cows. The reference test for hyperketonemia was defined as plasma concentration of beta-hydroxybutyrate (BHBA) >or=1,200 micromol/L. The weekly prevalence of hyperketonemia during the first 9 wk of lactation was, on average, 7.1%. Both BHBA and acetone in milk, determined by FTIR, had a higher sensitivity (80%) to detect hyperketonemia compared with milk fat to protein ratio (66%). Specificity was similar for the 3 diagnostic tests (71, 70, and 71%). In conclusion, FTIR predictions of BHBA or acetone in milk can detect cows with hyperketonemia in early lactation with a higher accuracy compared with the use of milk fat to protein ratio. Because of the high proportion of false-positive tests, there are concerns about the practical applicability of FTIR predictions of acetone, BHBA, and fat to protein ratio in milk to detect hyperketonemic cows.

Genomic breeding value estimation using genetic markers, inferred ancestral haplotypes, and the genomic relationship matrix

With the introduction of new single nucleotide polymorphism (SNP) chips of various densities, more and more genotype data sets will include animals genotyped for only a subset of the SNP. Imputation techniques based on unobserved ancestral haplotypes may be used to infer missing genotypes. These ancestral haplotypes may also be used in the genomic prediction model, instead of using the SNP. This may increase the reliability of predictions because the ancestral haplotype may capture more linkage disequilibrium with quantitative trait loci than SNP. The aim of this paper was to study whether using unobserved ancestral haplotypes in a genomic prediction model would provide more reliable genomic predictions than using SNP, and to determine how many loci in the genomic prediction model would be redundant. Genotypes of 8,960 bulls and cows for 39,557 SNP were analyzed with a hidden Markov model to associate each individual at each locus to 2 ancestral haplotypes. The number of ancestral haplotypes per locus was fixed at 10, 15, or 20. Subsequently, a validation study was performed in which the phenotypes of 3,251 progeny-tested bulls for 16 traits were used in a genomic prediction model to predict the estimated breeding values of at least 753 validation bulls. The squared correlation between genomic prediction and deregressed daughter performance estimated breeding value, when averaged across traits, was slightly higher when 15 or 20 ancestral haplotypes per locus were used in the prediction model instead of the SNP genotypes, whereas the prediction model using a genomic relationship matrix gave the lowest squared correlations. The number of redundant loci [i.e., loci that had less than 18 jumps (0.1%) from one ancestral haplotype to another ancestral haplotype at the next locus], was 18,793 (48%), which means that only 20,764 loci would need to be included in the genomic prediction model. This provides opportunities for greatly decreasing computer requirements of genomic evaluations with very large numbers of markers.