David Habier

Extension of the bayesian alphabet for genomic selection

BackgroundTwo Bayesian methods, BayesCπ and BayesDπ, were developed for genomic prediction to address the drawback of BayesA and BayesB regarding the impact of prior hyperparameters and treat the prior probability π that a SNP has zero effect as unknown. The methods were compared in terms of inference of the number of QTL and accuracy of genomic estimated breeding values (GEBVs), using simulated scenarios and real data from North American Holstein bulls.ResultsEstimates of π from BayesCπ, in contrast to BayesDπ, were sensitive to the number of simulated QTL and training data size, and provide information about genetic architecture. Milk yield and fat yield have QTL with larger effects than protein yield and somatic cell score. The drawback of BayesA and BayesB did not impair the accuracy of GEBVs. Accuracies of alternative Bayesian methods were similar. BayesA was a good choice for GEBV with the real data. Computing time was shorter for BayesCπ than for BayesDπ, and longest for our implementation of BayesA.ConclusionsCollectively, accounting for computing effort, uncertainty as to the number of QTL (which affects the GEBV accuracy of alternative methods), and fundamental interest in the number of QTL underlying quantitative traits, we believe that BayesCπ has merit for routine applications.

Genomic-BLUP Decoded: A Look into the Black Box of Genomic Prediction

Genomic best linear unbiased prediction (BLUP) is a statistical method that uses relationships between individuals calculated from single-nucleotide polymorphisms (SNPs) to capture relationships at quantitative trait loci (QTL). We show that genomic BLUP exploits not only linkage disequilibrium (LD) and additive-genetic relationships, but also cosegregation to capture relationships at QTL. Simulations were used to study the contributions of those types of information to accuracy of genomic estimated breeding values (GEBVs), their persistence over generations without retraining, and their effect on the correlation of GEBVs within families. We show that accuracy of GEBVs based on additive-genetic relationships can decline with increasing training data size and speculate that modeling polygenic effects via pedigree relationships jointly with genomic breeding values using Bayesian methods may prevent that decline. Cosegregation information from half sibs contributes little to accuracy of GEBVs in current dairy cattle breeding schemes but from full sibs it contributes considerably to accuracy within family in corn breeding. Cosegregation information also declines with increasing training data size, and its persistence over generations is lower than that of LD, suggesting the need to model LD and cosegregation explicitly. The correlation between GEBVs within families depends largely on additive-genetic relationship information, which is determined by the effective number of SNPs and training data size. As genomic BLUP cannot capture short-range LD information well, we recommend Bayesian methods with t-distributed priors.

Breeding value prediction for production traits in layer chickens using pedigree or genomic relationships in a reduced animal model

BackgroundGenomic selection involves breeding value estimation of selection candidates based on high-density SNP genotypes. To quantify the potential benefit of genomic selection, accuracies of estimated breeding values (EBV) obtained with different methods using pedigree or high-density SNP genotypes were evaluated and compared in a commercial layer chicken breeding line.MethodsThe following traits were analyzed: egg production, egg weight, egg color, shell strength, age at sexual maturity, body weight, albumen height, and yolk weight. Predictions appropriate for early or late selection were compared. A total of 2,708 birds were genotyped for 23,356 segregating SNP, including 1,563 females with records. Phenotypes on relatives without genotypes were incorporated in the analysis (in total 13,049 production records).The data were analyzed with a Reduced Animal Model using a relationship matrix based on pedigree data or on marker genotypes and with a Bayesian method using model averaging. Using a validation set that consisted of individuals from the generation following training, these methods were compared by correlating EBV with phenotypes corrected for fixed effects, selecting the top 30 individuals based on EBV and evaluating their mean phenotype, and by regressing phenotypes on EBV.ResultsUsing high-density SNP genotypes increased accuracies of EBV up to two-fold for selection at an early age and by up to 88% for selection at a later age. Accuracy increases at an early age can be mostly attributed to improved estimates of parental EBV for shell quality and egg production, while for other egg quality traits it is mostly due to improved estimates of Mendelian sampling effects. A relatively small number of markers was sufficient to explain most of the genetic variation for egg weight and body weight.

Genome-wide association analysis and genetic architecture of egg weight and egg uniformity in layer chickens.

The pioneering work by Professor Soller et al., among others, on the use of genetic markers to analyze quantitative traits has provided opportunities to discover their genetic architecture in livestock by identifying quantitative trait loci (QTL). The recent availability of high-density single nucleotide polymorphism (SNP) panels has advanced such studies by capitalizing on population-wide linkage disequilibrium at positions across the genome. In this study, genomic prediction model Bayes-B was used to identify genomic regions associated with the mean and standard deviation of egg weight at three ages in a commercial brown egg layer line. A total of 24,425 segregating SNPs were evaluated simultaneously using over 2900 genotyped individuals or families. The corresponding phenotypic records were represented as individual measurements or family means from full-sib progeny. A novel approach using the posterior distribution of window variances from the Monte Carlo Markov Chain samples was used to describe genetic architecture and to make statistical inferences about regions with the largest effects. A QTL region on chromosome 4 was found to explain a large proportion of the genetic variance for the mean (30%) and standard deviation (up to 16%) of the weight of eggs laid at specific ages. Additional regions with smaller effects on chromosomes 2, 5, 6, 8, 20, 23, 28 and Z showed suggestive associations with mean egg weight and a region on chromosome 13 with the standard deviation of egg weight at 26-28 weeks of age. The genetic architecture of the analyzed traits was characterized by a limited number of genes or genomic regions with large effects and many regions with small polygenic effects. The region on chromosome 4 can be used to improve both the mean and standard deviation of egg weight by marker-assisted selection.

Persistence of accuracy of genomic estimated breeding values over generations in layer chickens

BackgroundThe predictive ability of genomic estimated breeding values (GEBV) originates both from associations between high-density markers and QTL (Quantitative Trait Loci) and from pedigree information. Thus, GEBV are expected to provide more persistent accuracy over successive generations than breeding values estimated using pedigree-based methods. The objective of this study was to evaluate the accuracy of GEBV in a closed population of layer chickens and to quantify their persistence over five successive generations using marker or pedigree information.MethodsThe training data consisted of 16 traits and 777 genotyped animals from two generations of a brown-egg layer breeding line, 295 of which had individual phenotype records, while others had phenotypes on 2,738 non-genotyped relatives, or similar data accumulated over up to five generations. Validation data included phenotyped and genotyped birds from five subsequent generations (on average 306 birds/generation). Birds were genotyped for 23,356 segregating SNP. Animal models using genomic or pedigree relationship matrices and Bayesian model averaging methods were used for training analyses. Accuracy was evaluated as the correlation between EBV and phenotype in validation divided by the square root of trait heritability.ResultsPedigree relationships in outbred populations are reduced by 50% at each meiosis, therefore accuracy is expected to decrease by the square root of 0.5 every generation, as observed for pedigree-based EBV (Estimated Breeding Values). In contrast the GEBV accuracy was more persistent, although the drop in accuracy was substantial in the first generation. Traits that were considered to be influenced by fewer QTL and to have a higher heritability maintained a higher GEBV accuracy over generations. In conclusion, GEBV capture information beyond pedigree relationships, but retraining every generation is recommended for genomic selection in closed breeding populations.

Extent and Consistency of Linkage Disequilibrium and Identification of DNA Markers for Production and Egg Quality Traits in Commercial Layer Chicken Populations

BackgroundThe genome sequence and a high-density SNP map are now available for the chicken and can be used to identify genetic markers for use in marker-assisted selection (MAS). Effective MAS requires high linkage disequilibrium (LD) between markers and quantitative trait loci (QTL), and sustained marker-QTL LD over generations. This study used data from a 3,000 SNP panel to assess the level and consistency of LD between single nucleotide polymorphisms (SNPs) over consecutive years in two egg-layer chicken lines, and analyzed one line by two methods (SNP-wise association and genome-wise Bayesian analysis) to identify markers associated with egg-quality and egg-production phenotypes.ResultsThe LD between markers pairs was high at short distances (r2 > 0.2 at < 2 Mb) and remained high after one generation (correlations of 0.80 to 0.92 at < 5 Mb) in both lines. Single- and 3-SNP regression analyses using a mixed model with SNP as fixed effect resulted in 159 and 76 significant tests (P < 0.01), respectively, across 12 traits. A Bayesian analysis called BayesB, that fits all SNPs simultaneously as random effects and uses model averaging procedures, identified 33 SNPs that were included in the model >20% of the time (φ > 0.2) and an additional ten 3-SNP windows that had a sum of φ greater than 0.35. Generally, SNPs included in the Bayesian model also had a small P-value in the 1-SNP analyses.ConclusionHigh LD correlations between markers at short distances across two generations indicate that such markers will retain high LD with linked QTL and be effective for MAS. The different association analysis methods used provided consistent results. Multiple single SNPs and 3-SNP windows were significantly associated with egg-related traits, providing genomic positions of QTL that can be useful for both MAS and to identify causal mutations.

Genomic breeding value prediction and QTL mapping of QTLMAS2010 data using Bayesian Methods

BackgroundBayesian methods allow prediction of genomic breeding values (GEBVs) using high-density single nucleotide polymorphisms (SNPs) covering the whole genome with effective shrinkage of SNP effects using appropriate priors. In this study we applied a modification of the well-known BayesA and BayesB methods to estimate the proportion of SNPs with zero effects (π) and a common variance for non-zero effects. The method, termed BayesCπ, was used to predict the GEBVs of the last generation of the QTLMAS2010 data. The accuracy of GEBVs from various methods was estimated by the correlation with phenotypes in the last generation. The methods were BayesCPi and BayesB with different π values, both with and without polygenic effects, and best linear unbiased prediction using an animal model with a genomic or numerator relationship matrix. Positions of quantitative trait loci (QTLs) were identified based on the variances of GEBVs for windows of 10 consecutive SNPs. We also proposed a novel approach to set significance thresholds for claiming QTL in this specific case by using pedigree-based simulation of genotypes. All analyses were focused on detecting and evaluating QTL with additive effects.ResultsThe accuracy of GEBVs was highest for BayesCπ, but the accuracy of BayesB with π equal to 0.99 was similar to that of BayesCπ. The accuracy of BayesB dropped with a decrease in π. Including polygenic effects into the model only had marginal effects on accuracy and bias of predictions. The number of QTL identified was 15 when based on a stringent 10% chromosome-wise threshold and increased to 21 when a 20% chromosome-wise threshold was used.ConclusionsThe BayesCπ method without polygenic effects was identified to be the best method for the QTLMAS2010 dataset, because it had highest accuracy and least bias. The significance criterion based on variance of 10-SNP windows allowed detection of more than half of the QTL, with few false positives.

Accuracy of genomic prediction using an evenly spaced, low-density single nucleotide polymorphism panel in broiler chickens

One approach for cost-effective implementation of genomic selection is to genotype training individuals with a high-density (HD) panel and selection candidates with an evenly spaced, low-density (ELD) panel. The purpose of this study was to evaluate the extent to which the ELD approach reduces the accuracy of genomic estimated breeding values (GEBV) in a broiler line, in which 1,091 breeders from 3 generations were used for training and 160 progeny of the third generation for validation. All birds were genotyped with an Illumina Infinium platform HD panel that included 20,541 segregating markers. Two subsets of HD markers, with 377 (ELD-1) or 766 (ELD-2) markers, were selected as ELD panels. The ELD-1 panel was genotyped using KBiosciences KASPar SNP genotyping chemistry, whereas the ELD-2 panel was simulated by adding markers from the HD panel to the ELD-1 panel. The training data set was used for 2 traits: BW at 35 d on both sexes and hen house production (HHP) between wk 28 and 54. Methods Bayes-A, -B, -C and genomic best linear unbiased prediction were used to estimate HD-marker effects. Two scenarios were used: (1) the 160 progeny were ELD-genotyped, and (2) the 160 progeny and their dams (117 birds) were ELD-genotyped. The missing HD genotypes in ELD-genotyped birds were imputed by a Gibbs sampler, capitalizing on linkage within families. In scenario (1), the correlation of GEBV for BW (HHP) of the 160 progeny based on observed HD versus imputed genotypes was greater than 0.94 (0.98) with the ELD-1 panel and greater than 0.97 (0.99) with the ELD-2 panel. In scenario (2), the correlation of GEBV for BW (HHP) was greater than 0.92 (0.96) with the ELD-1 panel and greater than 0.95 (0.98) with the ELD-2 panel. Hence, in a pedigreed population, genomic selection can be implemented by genotyping selection candidates with about 400 ELD markers with less than 6% loss in accuracy. This leads to substantial savings in genotyping costs, with little sacrifice in accuracy.

A Two-Stage Approximation for Analysis of Mixture Genetic Models in Large Pedigrees

Information from cosegregation of marker and QTL alleles, in addition to linkage disequilibrium (LD), can improve genomic selection. Variance components linear models have been proposed for this purpose, but accommodating dominance and epistasis is not straightforward with them. A full-Bayesian analysis of a mixture genetic model is favorable in this respect, but is computationally infeasible for whole-genome analyses. Thus, we propose an approximate two-step approach that neglects information from trait phenotypes in inferring ordered genotypes and segregation indicators of markers. Quantitative trait loci (QTL) fine-mapping scenarios, using high-density markers and pedigrees of five generations without genotyped females, were simulated to test this strategy against an exact full-Bayesian approach. The latter performed better in estimating QTL genotypes, but precision of QTL location and accuracy of genomic breeding values (GEBVs) did not differ for the two methods at realistically low LD. If, however, LD was higher, the exact approach resulted in a slightly higher accuracy of GEBVs. In conclusion, the two-step approach makes mixture genetic models computationally feasible for high-density markers and large pedigrees. Furthermore, markers need to be sampled only once and results can be used for the analysis of all traits. Further research is needed to evaluate the two-step approach for complex pedigrees and to analyze alternative strategies for modeling LD between QTL and markers.

Breeding Value Prediction for Production Traits in Layers Using High-density SNP Markers

Introduction Through the application of genomic or whole-genome selection (Meuwissen et al. 2001), marker information from high-density SNP genotyping can improve accuracy of selection at young ages, shorten generation intervals and provide better control of inbreeding, which should lead to higher genetic gain per time unit. Multiple simulation studies have been conducted showing that the benefits of the technology depend on heritability, number and effects of QTL, population structure, size of the training data set, and other factors (Goddard, 2009). There are however few studies on real data. The objective of this study was to evaluate the accuracy of breeding values estimated using high-density SNP genotypes in predicting the next generation in a commercial layer breeding line, and to compare the accuracy of different methods of breeding value estimation.