La mise en place de la sélection génomique rend possible l’inclusion de nouveaux caractères dans les objectifs de sélection, en profitant des marges offertes par l’augmentation du progrès génétique sur les caractères historiques. En parallèle, les attentes des éleveurs, de la société et des filières orientent la sélection génétique vers des objectifs plus larges et complexes. Deux groupes de nouveaux caractères ont été analysés dans le cadre de l’amélioration génétique des bovins laitiers : les caractères de carcasse des jeunes bovins de races mixtes et les caractères de résistance aux lésions podales en race Holstein, afin de préparer la mise en place d’une sélection génétique et génomique de ces caractères. Pour ces deux caractères, les modèles d’évaluation polygénique les plus pertinents ont été développés et les paramètres génétiques correspondant ont été estimés. Ces derniers révèlent que la sélection génétique des caractères de carcasse des jeunes bovins sera relativement aisée, alors que celle de la résistance aux lésions podales sera plus difficile, mais possible, du fait de l’existence d’une variabilité génétique. Ils mettent également en évidence qu’un progrès génétique sur les caractères de carcasse des jeunes bovins n’est pas génétiquement opposé à un progrès génétique sur la production laitière. Ils mettent enfin en évidence deux groupes de lésions podales génétiquement distinctes. Par ailleurs, plusieurs stratégies pour prendre en compte la nonexhaustivité des données de lésions podales ont été testées. Plusieurs approches d’évaluation ont été comparées. Pour les deux groupes de caractères, l’approche dite « Single-Step Genomic BLUP » s’est révélée la plus prometteuse, même si les autres approches génomiques « en deux étapes », donnaient des résultats aux caractéristiques relativement proches. Ces études ont mené à la mise en place d’évaluations génétiques et génomiques en routine pour les deux groupes de caractères, pour lesquelles les méthodes d’évaluation classiques ont été retenues par souci de cohérence avec les évaluations actuelles, tout en démontrant l’intérêt d’une évolution vers l’approche en une étape à moyen terme. Les principales questions soulevées et les grandes étapes identifiées lors des études de ces deux caractères ont été rassemblées au sein d’une ébauche de méthodologie pour le développement d’évaluations génétiques pour de nouveaux caractères. / The implementation of genomic selection makes possible the inclusion of new traits in breeding goals, by taking advantage of the opportunities coming from the increased genetic trend on traits currently under selection. Breeders, breeding companies and society all have changing expectations regarding genetic selection. Two groups of new traits were analyzed in the context of genetic improvement of dairy cattle: carcass traits of young bulls in dual-purpose breeds and claw health traits in Holstein, in order to prepare the implementation of genetic and genomic selection on these traits. For both sets of traits, suitable genetic evaluation models were developed and genetic parameters were estimated. Genetic parameters reveal that genetic selection of carcass traits of young bulls appears to be fairly easy and that selection of claw health traits is going to be more difficult, but possible, given the existing genetic variation. They also highlight that there is no strong negative genetic correlation between carcass traits of young bulls and dairy production traits. Finally, they reveal that there are two genetically distinct groups of claw health traits. Several strategies to account for non-exhaustive recording of cows for trimming were tested. Several evaluation approaches were compared. For both sets of traits, Single-Step Genomic BLUP was the most promising approach, although other (two-step) genomic approaches allowed for relatively similar accuracies and control of bias. These studies led to the implementation of routine genetic and genomic evaluations for both sets of traits, for which a usual two-step genomic approach was preferred over Single-Step Genomic BLUP for consistency with the current evaluation of other traits. However these two examples illustrate the benefit of implementing routine Single-Step Genomic BLUP evaluations. The main questions and principal steps identified in these studies were gathered into tentative guidelines for the development of genetic evaluations for new traits.

I sistemi di valutazione genetica nel mondo sono in rapido sviluppo. Attualmente, i programmi di selezione “tradizionale” basati su fenotipi e rapporti di parentela tra gli animali vengono integrati, e nel futuro potrebbero essere sostituiti, dalle informazioni molecolari. In questo periodo di transizione, questa tesi riguarda ricerche su entrambi i tipi di valutazioni: dall’accertamento sull’accuratezza degli indici genetici internazionali (tradizionali), allo studio di metodi statistici utilizzati per integrare informazioni genomiche nella selezione (selezione genomica). Tre capitoli valutano gli approcci per stimare i valori genetici dai dati genomici riducendo il numero di variabili indipendenti. In modo particolare, la correzione di Bonferroni e il test di permutazioni con regressione a marcatori singoli (Capitolo III), analisi delle componenti principali con BLUP (Capitolo IV) e indice Fst tra razze con BayesA (Capitolo VI). Inoltre, il Capitolo V analizza l’accuratezza dei valori genomici con BLUP, BayesA e Bayesian LASSO includendo tutte le variabili disponibili. I risultati di questa tesi indicano che il progresso genetico atteso dall’analisi dei dati simulati può effettivamente essere ottenuto, anche se ulteriori ricerche sono necessarie per ottimizzare l’utilizzo delle informazioni molecolari in modo da ottimizzare i risultati per tutti i caratteri sotto selezione. / Genetic evaluation systems are in rapid development worldwide. In most countries, “traditional” breeding programs based on phenotypes and relationships between animals are currently being integrated and in the future might be replaced by the introduction of molecular information. This thesis stands in this transition period, therefore it covers research on both types of genetic evaluations: from the assessment of the accuracy of (traditional) international genetic evaluations to the study of statistical methods used to integrate genomic information into breeding (genomic selection). Three chapters investigate and evaluate approaches for the estimation of genetic values from genomic data reducing the number of independent variables. In particular, Bonferroni correction and Permutation test combined with single marker regression (Chapter III), principal component analysis combined with BLUP (Chapter IV) and Fst across breeds combined with BayesA (Chapter VI). In addition, Chapter V analyzes the accuracy of direct genomic values with BLUP, BayesA and Bayesian LASSO including all available variables. The results of this thesis indicate that the genetic gains expected from the analysis of simulated data can be obtained on real data. Still, further research is needed to optimize the use of genome-wide information and obtain the best possible estimates for all traits under selection.

La santé du constitue une préoccupation croissante pour les éleveurs, propriétaires, et le grand public, plusieurs rapports ayant récemment souligné les potentiels impacts négatifs des pratiques d'élevage sur la santé des chiens de race (APGAW 2009, Nicholas 2011), au travers de la diffusion d’affections héréditaires par exemple. Ainsi, l’OFA (Orthopedic Foundation for Animals, http://www.offa.org) considère que la dysplasie de la hanche affecte au moins 163 races de chiens, avec des prévalences allant de 1,2 à 72,1%. La mise en œuvre de stratégie d'élevage afin de réduire l'incidence des maladies héréditaires et leur impact sur le bien-être constitue une priorité pour les éleveurs et les organisations d'élevage. L'efficacité de ces stratégies dépend toutefois fortement de facteurs tels que leur déterminisme génétique, la disponibilité de diagnostics cliniques ou génétiques efficaces, ainsi que les conditions spécifiques au contexte (la prévalence, la démographie, l'existence d'autres affections, la coopération des éleveurs ...). Par exemple, il a été montré que pour une maladie monogénique récessive, à fréquence égale, l'impact d’une stratégie sur la variabilité génétique sera extrêmement différent en fonction de la race (Leroy et Rognon 2012). Il est important de souligner également que le contexte et le cadre réglementaire de l'élevage varient beaucoup en fonction des pays. A titre d'exemple, en Suède, la proportion importante d'animaux de compagnie assurés (environ 50%) permet la mise en place d’enquêtes sur la santé des chiens à grande échelle (Bonnett et al., 2005), facilitant l'identification des affections impactant le bien-être. En fonction des pays, des mesures différentes de luttes contre les affections héréditaires ont pu être mises en place, pouvant aller de l’incitation à utiliser des reproducteurs sains, à l’interdiction de reproduction pour des individus atteints d’affection problématiques. Dans le cas de dysplasie de la hanche, un système d'évaluation génétique a été mis en œuvre dans certains pays (Allemagne, Suède, Royaume-Uni) pour quelques races, alors que dans certains autres pays, il est encore en cours de développement. Notons qu’un projet préliminaire à la thèse sera mise en place à l’échelle des kennels clubs nordiques (KNU) pour s’intéresser à la valeur ajoutée des échanges internationaux de données généalogiques et de santé. / Dog health constitutes a major concern for breeders, owners, as well as the general public, all the more since several study and reports have recently underlined potential impacts of breeding practices on dog health and fitness (APGAW 2009, Nicholas 2011). According to Online Mendelian Inheritance in Animals (OMIA, omia.angis.org.au) more than 586 disorders/traits have been reported in dogs, with various prevalence and consequences for canine health (Collins et al. 2011, Nicholas et al. 2011). As an exemple, Orthopedic Foundation for Animals (OFA 2011, http://www.offa.org) consider that Hip Dysplasia, a polygenetic trait affected by environmental factors, with variable impact on welfare, affects at least 163 dog breeds, with prevalence ranging from 1.2 to 72.1%. Implementation of breeding plans in order to reduce incidence of inherited disorders and their impact on welfare should be a priority for breeders and breeding organizations. Efficiency of such strategies is however highly dependant on several factors such as inheritance pattern, availability of efficient clinical/genetic test, and specific context conditions (prevalence, demography, existence of other disorders, cooperation of breeders…). For instance, it has been showed that for a monogenic recessive disorder with the same frequency, impact of a given strategy on genetic diversity will be completely different depending on the breed (Leroy and Rognon 2012). It is also important to underline that breeding context and breeding rules are very different according to countries. As an exemple, in Sweden the large proportion of pets insured (about 50%) allows the settlement of large surveys on dog health (Bonnett et al. 2005), leading to the identification of disorders critical to breed welfare. Depending on countries, the control of inherited disorders is implemented through various measures, from breeding recommandations to mating ban. In the case of hip dysplasia, a genetic evaluation system has been implemented in some countries (Germany, Sweden, UK) for a few breeds, while in some other countries, it is still under development. The fact that for many breeds there is an exchange of breeding animals between several countries with different breeding policies constitutes also a critical point to be taken into account, when settling a breeding strategy. Moreover it has been showed that efficiency of genetic evaluation for a polygenic trait such a hip dysplasia could be improved by joint evaluation between different countries (Fikse et al. 2012). For this purpose, a preliminary project, starting 2013 in Sweden, will investigate the interest of exchanging pedigree and health data within the framework of Nordic Kennel Union. The aim of this project is to provide operational tools to improve breed health in an international context, concerning both genetic evaluation and implementation of breeding policies.

<p>Breeding for improved animal welfare is paramount for increasing the long-term sustainability of the animal food industry. In this context, the main objectives of this dissertation were to understand the genetic and genomic background of welfare indicators in livestock and evaluate the feasibility of single-step Genomic Best Linear Unbiased Prediction (ssGBLUP) for performing genomic selection in beef cattle. This dissertation includes five studies. First, we aimed to test and identify an optimal ssGBLUP scenario for crossbreeding schemes. We simulated multiple populations differing based on the genetic background of the trait, and then we tested alternative models, such as multiple-trait weighted ssGBLUP. Even though more elaborated scenarios were evaluated, a single-trait ssGBLUP approach was recommended when genetic correlation across populations were higher than 0.70. The goal of the second study was to identify genomic regions controlling behavior traits that are conserved across livestock species. We systematically reviewed genomic regions associated with behavioral indicators in beef and dairy cattle, pigs, and sheep. The genomic regions identified in this study were located in genes previously reported controlling human behavioral, neural, and mental disorders. In the third study we used a large dataset (675,678 records) from North American Angus cattle to investigate the genetic background of temperament, a behavioral indicator, recorded on one-year-old calves, and provide the models and protocols for implementing genomic selection. We reported a heritability estimate equal to 0.38 for yearling temperament, and it was, in general, genetically favorably correlated with other productivity and fertility traits. Candidate genomic regions controlling yearling temperament were also identified. The fourth study was based on temperament recorded on North American Angus cows from 2 to 15 years of age (797,187 records). The goal was to understand the genetic and genomic background of temperament across the animal’s lifetime. By fitting a random regression model, we observed that temperament is highly genetically correlated across time. However, animals have differential learning and behavioral plasticity (LBP; changes of the phenotype overtime), although the LBP heritability is low. In our last study we evaluated foot scores (foot angle, FA; and claw set, CS) in American (US) and Australian (AU) Angus cattle aiming to assess the genetic and genomic background of foot scores and investigate the feasibility of performing an across-country genomic evaluation (~1.15 million animals genotyped). Foot scores are heritable (heritability from 0.22 to 0.27), and genotype-by-environment interaction was observed between US and AU Angus populations (genetic correlation equal to 0.61 for FA and 0.76 for CS). An across-country genomic prediction outperformed within-country evaluations in terms of predictivity ability (bias, dispersion, and validation accuracy) and theoretical accuracies. We have also identified genes associated with FA and CS previously reported in human’s bone structure and repair mechanism. In conclusion, this dissertation presents a comprehensive genetic and genomic characterization of welfare indicators (temperament and foot scores) in (inter)national livestock populations. </p>