Addie Vereijken

Hybridization and/or introgression play a key role in the evolutionary history of animal species. It is commonly observed in several orders in wild birds. The domestic chicken Gallus gallus domesticus is the commonest livestock species exploited for the benefit of mankind. The origin of its diversity remains unsettled. Here, we report a genome-wide analyses for signatures of introgression within domestic village chicken. We first established the genome-wide phylogeny and divergence time across the genus Gallus , showing the sister relationships between Grey junglefowl G. sonneratii and Ceylon junglefowl G. lafayettii and that the Green junglefowl is the first diverging lineage within the genus Gallus . Then, by analysing the whole-genome sequences of geographically diverse chicken populations, we reveal extensive bidirectional introgression between Grey junglefowl and domestic chicken and to a much less extent with Ceylon junglefowl. A single case of Green junglefowl G. varius introgression was identified. These introgressed regions include biological functions related to the control of gene expression. Our results show that while the Red junglefowl is the main ancestral species, introgressive hybridization episodes have impacted the genome and contributed to the diversity of domestic chicken, although likely at different level across its geographic range.

Cannibalism is an important welfare problem in the layer industry. Cannibalism is a social behavior where individual survival is affected by direct genetic effects (DGE) and indirect genetic effects (IGE). Previous studies analysed repeated binomial survival, instead of survival time, which improved accuracies of breeding value predictions. Our study aimed at identifying SNPs associated with DGE and IGE for survival time, and comparing results from models that analyse survival time and repeated binomial survival. Survival data of three layer crosses (W1 * WA, W1 * WB, and W1 * WC) were used. Each individual had one survival time record and 13 monthly survival (0/1) records. Approximately 30,000 single nucleotide polymorphisms (SNPs) were included in the genome-wide association study (GWAS), using a linear mixed model for survival time, a linear mixed model and a generalized linear mixed model for repeated binomial survival (0/1). Backwards elimination was used to determine phenotypic and genetic variance explained by SNPs. The same quantitative trait loci were identified with all models. A SNP associated with DGE was found in cross W1 * WA, with an allele substitution effect of 22 days. This SNP explained 3% of the phenotypic variance, and 36% of the total genetic variance. Four SNPs associated with DGE were found in cross W1 * WB, with effects ranging from 16 to 35 days. These SNPs explained 1 to 6% of the phenotypic variance and 9 to 44% of the total genetic variance. Our results suggest a link of DGE and IGE for survival time in layers with the gamma-aminobutyric acid (GABA) system, since a SNP located near a gene for a GABA receptor was associated with DGE and with IGE (not significant). This is one of the first large studies investigating the genetic architecture of a socially-affected trait. The power to detect SNP associations was relatively low and thus we expect that many effects on DGE and IGE remained undetected. Yet, GWAS results revealed SNPs with large DGE and a link of DGE and IGE for survival time in layers with the GABAergic system, which supports existing evidence for the involvement of GABA in the development of abnormal behaviors.

Sex-linked slow (SF) and fast (FF) feathering rates at hatch have been widely used in poultry breeding for autosexing at hatch. In chicken, the sex-linked K (SF) and k+ (FF) alleles are responsible for the feathering rate phenotype. Allele K is dominant and a partial duplication of the prolactin receptor gene has been identified as the causal mutation. Interestingly, some domesticated turkey lines exhibit similar slow- and fast-feathering phenotypes, but the underlying genetic components and causal mutation have never been investigated. In this study, our aim was to investigate the molecular basis of feathering rate at hatch in domestic turkey. We performed a sequence-based case–control association study and detected a genomic region on chromosome Z, which is statistically associated with rate of feathering at hatch in turkey. We identified a 5-bp frameshift deletion in the prolactin receptor (PRLR) gene that is responsible for slow feathering at hatch. All female cases (SF turkeys) were hemizygous for this deletion, while 188 controls (FF turkeys) were hemizygous or homozygous for the reference allele. This frameshift mutation introduces a premature stop codon and six novel amino acids (AA), which results in a truncated PRLR protein that lacks 98 C-terminal AA. We present the causal mutation for feathering rate in turkey that causes a partial C-terminal loss of the prolactin receptor, and this truncated PRLR protein is strikingly similar to the protein encoded by the slow feathering K allele in chicken.

Mortality due to cannibalism causes both economic and welfare problems in laying hens. To limit mortality due to cannibalism, laying hens are often beak-trimmed, which is undesirable for animal welfare reasons. Genetic selection is an alternative strategy to increase survival and is more efficient by taking heritable variation that originates from social interactions into account, which are modelled as the so-called indirect genetic effects (IGE). Despite the considerable heritable variation in survival time due to IGE, genetic improvement of survival time in laying hens is still challenging because the detected heritable variation of the trait with IGE is still limited, ranging from 0.06 to 0.26, and individuals that are still alive at the end of the recording period are censored. Furthermore, survival time records are available late in life and only on females. To cope with these challenges, we tested the hypothesis that genomic prediction increases the accuracy of estimated breeding values (EBV) compared to parental average EBV, and increases response to selection for survival time compared to a traditional breeding scheme. We tested this hypothesis in two lines of brown layers with intact beaks, which show cannibalism, and also the hypothesis that the rate of inbreeding per year is lower for genomic selection than for the traditional breeding scheme. The standard deviation of genomic prediction EBV for survival time was around 22 days for both lines, indicating good prospects for selection against mortality in laying hens with intact beaks. Genomic prediction increased the accuracy of the EBV by 35 and 32 % compared to the parent average EBV for the two lines. At the current reference population size, predicted response to selection was 91 % higher when using genomic selection than with the traditional breeding scheme, as a result of a shorter generation interval in males and greater accuracy of selection in females. The predicted rate of inbreeding per generation with truncation selection was substantially lower for genomic selection than for the traditional breeding scheme for both lines. Genomic selection for socially affected traits is a promising tool for the improvement of survival time in laying hens with intact beaks.

Genotype imputation has become a standard practice in modern genetic research to increase genome coverage and improve the accuracy of genomic selection (GS) and genome-wide association studies (GWAS). We assessed accuracies of imputing 60K genotype data from lower density single nucleotide polymorphism (SNP) panels using a small set of the most common sires in a population of 2140 white layer chickens. Several factors affecting imputation accuracy were investigated, including the size of the reference population, the level of the relationship between the reference and validation populations, and minor allele frequency (MAF) of the SNP being imputed. The accuracy of imputation was assessed with different scenarios using 22 and 62 carefully selected reference animals (Ref22 and Ref62). Animal-specific imputation accuracy corrected for gene content was moderate on average (~ 0.80) in most scenarios and low in the 3K to 60K scenario. Maximum average accuracies were 0.90 and 0.93 for the most favourable scenario for Ref22 and Ref62 respectively, when SNPs were masked independent of their MAF. SNPs with low MAF were more difficult to impute, and the larger reference population considerably improved the imputation accuracy for these rare SNPs. When Ref22 was used for imputation, the average imputation accuracy decreased by 0.04 when validation population was two instead of one generation away from the reference and increased again by 0.05 when validation was three generations away. Selecting the reference animals from the most common sires, compared with random animals from the population, considerably improved imputation accuracy for low MAF SNPs, but gave only limited improvement for other MAF classes. The allelic R2 measure from Beagle software was found to be a good predictor of imputation reliability (correlation ~ 0.8) when the density of validation panel was very low (3K) and the MAF of the SNP and the size of the reference population were not extremely small. Even with a very small number of animals in the reference population, reasonable accuracy of imputation can be achieved. Selecting a set of the most common sires, rather than selecting random animals for the reference population, improves the imputation accuracy of rare alleles, which may be a benefit when imputing with whole genome re-sequencing data.

Most studies on genomic prediction with reference populations that include multiple lines or breeds have used linear models. Data heterogeneity due to using multiple populations may conflict with model assumptions used in linear regression methods. In an attempt to alleviate potential discrepancies between assumptions of linear models and multi-population data, two types of alternative models were used: (1) a multi-trait genomic best linear unbiased prediction (GBLUP) model that modelled trait by line combinations as separate but correlated traits and (2) non-linear models based on kernel learning. These models were compared to conventional linear models for genomic prediction for two lines of brown layer hens (B1 and B2) and one line of white hens (W1). The three lines each had 1004 to 1023 training and 238 to 240 validation animals. Prediction accuracy was evaluated by estimating the correlation between observed phenotypes and predicted breeding values. When the training dataset included only data from the evaluated line, non-linear models yielded at best a similar accuracy as linear models. In some cases, when adding a distantly related line, the linear models showed a slight decrease in performance, while non-linear models generally showed no change in accuracy. When only information from a closely related line was used for training, linear models and non-linear radial basis function (RBF) kernel models performed similarly. The multi-trait GBLUP model took advantage of the estimated genetic correlations between the lines. Combining linear and non-linear models improved the accuracy of multi-line genomic prediction. Linear models and non-linear RBF models performed very similarly for genomic prediction, despite the expectation that non-linear models could deal better with the heterogeneous multi-population data. This heterogeneity of the data can be overcome by modelling trait by line combinations as separate but correlated traits, which avoids the occasional occurrence of large negative accuracies when the evaluated line was not included in the training dataset. Furthermore, when using a multi-line training dataset, non-linear models provided information on the genotype data that was complementary to the linear models, which indicates that the underlying data distributions of the three studied lines were indeed heterogeneous.

The prediction accuracy of several linear genomic prediction models, which have previously been used for within-line genomic prediction, was evaluated for multi-line genomic prediction. Compared to a conventional BLUP (best linear unbiased prediction) model using pedigree data, we evaluated the following genomic prediction models: genome-enabled BLUP (GBLUP), ridge regression BLUP (RRBLUP), principal component analysis followed by ridge regression (RRPCA), BayesC and Bayesian stochastic search variable selection. Prediction accuracy was measured as the correlation between predicted breeding values and observed phenotypes divided by the square root of the heritability. The data used concerned laying hens with phenotypes for number of eggs in the first production period and known genotypes. The hens were from two closely-related brown layer lines (B1 and B2), and a third distantly-related white layer line (W1). Lines had 1004 to 1023 training animals and 238 to 240 validation animals. Training datasets consisted of animals of either single lines, or a combination of two or all three lines, and had 30 508 to 45 974 segregating single nucleotide polymorphisms. Genomic prediction models yielded 0.13 to 0.16 higher accuracies than pedigree-based BLUP. When excluding the line itself from the training dataset, genomic predictions were generally inaccurate. Use of multiple lines marginally improved prediction accuracy for B2 but did not affect or slightly decreased prediction accuracy for B1 and W1. Differences between models were generally small except for RRPCA which gave considerably higher accuracies for B2. Correlations between genomic predictions from different methods were higher than 0.96 for W1 and higher than 0.88 for B1 and B2. The greater differences between methods for B1 and B2 were probably due to the lower accuracy of predictions for B1 (~0.45) and B2 (~0.40) compared to W1 (~0.76). Multi-line genomic prediction did not affect or slightly improved prediction accuracy for closely-related lines. For distantly-related lines, multi-line genomic prediction yielded similar or slightly lower accuracies than single-line genomic prediction. Bayesian variable selection and GBLUP generally gave similar accuracies. Overall, RRPCA yielded the greatest accuracies for two lines, suggesting that using PCA helps to alleviate the “n≪p” problem in genomic prediction.

In livestock species like the chicken, high throughput single nucleotide polymorphism (SNP) genotyping assays are increasingly being used for whole genome association studies and as a tool in breeding (referred to as genomic selection). To be of value in a wide variety of breeds and populations, the success rate of the SNP genotyping assay, the distribution of the SNP across the genome and the minor allele frequencies (MAF) of the SNPs used are extremely important.

Variation within individual genomes ranges from single nucleotide polymorphisms (SNPs) to kilobase, and even megabase, sized structural variants (SVs), such as deletions, insertions, inversions, and more complex rearrangements. Although much is known about the extent of SVs in humans and mice, species in which they exert significant effects on phenotypes, very little is known about the extent of SVs in the 2.5-times smaller and less repetitive genome of the chicken. We identified hundreds of shared and divergent SVs in four commercial chicken lines relative to the reference chicken genome. The majority of SVs were found in intronic and intergenic regions, and we also found SVs in the coding regions. To identify the SVs, we combined high-throughput short read paired-end sequencing of genomic reduced representation libraries (RRLs) of pooled samples from 25 individuals and computational mapping of DNA sequences from a reference genome. We provide a first glimpse of the high abundance of small structural genomic variations in the chicken. Extrapolating our results, we estimate that there are thousands of rearrangements in the chicken genome, the majority of which are located in non-coding regions. We observed that structural variation contributes to genetic differentiation among current domesticated chicken breeds and the Red Jungle Fowl. We expect that, because of their high abundance, SVs might explain phenotypic differences and play a role in the evolution of the chicken genome. Finally, our study exemplifies an efficient and cost-effective approach for identifying structural variation in sequenced genomes.

The turkey (Meleagris gallopavo) is an important agricultural species and is the second largest contributor to the world's poultry meat production. Demand of turkey meat is increasing very rapidly. Genetic markers linked to genes affecting quantitative traits can increase the selection response of animal breeding programs. The use of these molecular markers for the identification of quantitative trait loci, and subsequently fine-mapping of quantitative trait loci regions, allows for pinpointing of genes that underlie such economically important traits.