Cet article présente l’histoire et le bilan de la sélection génomique bovine en France, avec ses étapes de démarrage dans les grandes races laitières, son extension progressive à toutes les races, le développement du génotypage permis par la diminution de son coût, l’intégration de nouveaux caractères et la redéfinition des objectifs de sélection avec davantage d’importance donnée aux caractères fonctionnels. Il présente les évolutions méthodologiques réalisées au cours du temps, concernant le génotypage, les populations de référence et les méthodes d’évaluation génomique. Les résultats acquis sont présentés, en termes d’évolution génétique sur les différents caractères – toujours favorable – d’évolution de la consanguinité – très variable entre races – et de maîtrise des anomalies génétiques. Une forte synergie a été construite entre recherche et sélection : toutes les évolutions ont été suscitées par la recherche, mais d’un autre côté, les données nouvelles engendrées par la sélection ont nourri la recherche. Enfin, l’article décrit succinctement les fortes évolutions organisationnelles qui ont été pour partie induites par la sélection génomique.

Shifts in the dairy cattle industry have placed greater emphasis on breeding healthier animals. While adult cow health-related traits have historically been more studied than calf health-related traits, largely due to greater data availability, the significant impact of the latter on dairy production, especially calf survival, has gained increased attention. Several countries have already assessed the feasibility of implementing a genomic evaluation for this trait, whereas in France, only a genomic evaluation for perinatal mortality currently exists in dairy cattle breeds. Hence, implementing a routine genomic evaluation may help improve calf survival. In this context, this study first aimed at estimating genetic parameters for calf survival across 2 key subperiods (between d 3-30 and between d 31-365 where mortality is mainly associated with gastrointestinal diseases and with respiratory diseases, respectively) and across the entire calf survival period (d 3-365) in 9 French dairy cattle breeds. Second, we investigated whether calf survival should be divided into 2 subperiods to more precisely discriminate the causes of mortality or be considered as a single period. Third, genetic trends for calf survival traits were explored. Accordingly, linear animal models were used to estimate heritability and genetic correlations between calf survival traits. The analyses included up to 12 million pedigree records and between 4,750 and 8,296,055 records collected from 2010 to 2024, depending on the breed. Estimated breeding values of bulls obtained from genomic evaluations for calf survival were used to assess genetic trends. The 3 calf survival traits showed low heritabilities (h2 = 0.006-0.031) but displayed large genetic standard deviations (SD) making genomic selection for these traits possible. High genetic correlations were found between the calf survival subperiods, with most estimates near 0.80, indicating close genetic backgrounds. To better discriminate the causes of mortality, we propose combining both d3-30 and d31-365 in an index synthesis. Genetic trends were favorable after 2010 in the 3 major breeds (Holstein, Montbéliarde, and Normande), but no change was observed in the others. Future work will estimate genetic correlations between calf survival traits and traits under selection, thereby paving the way for including calf survival in French dairy cattle breeds' breeding programs to help decrease juvenile mortality.

Genomic selection (GS) has revolutionized animal breeding and accelerated genetic gains in breeding programs. Although GS has become common in cosmopolitan dairy cattle breeds, its implementation in local breeds has begun only more recently, or is still in progress. However, the introduction of GS in some cosmopolitan breeds has also been associated with increased inbreeding rates, raising concerns about the potential effects of GS on genetic diversity in smaller or local breeds. Our aim was to investigate the impact of GS on genetic diversity in 5 (small) local cattle breeds from 3 European countries: Meuse Rhine Yssel (MRY; from the Netherlands); Norwegian Red (NRC; from Norway); and Abondance (ABO), Tarentaise (TAR), and Vosgienne (VOS; from France). We investigated changes in population demographic structure, as well as trends and rates of kinship and inbreeding, using both pedigree- and genomic-based measures. The population size varied depending on the breed, with Vosgienne being the smallest and NRC being the largest. Single nucleotide polymorphism genotypes were available for 4,645 MRY, 193,489 NRC, 16,387 ABO, 8,578 TAR, and 4,472 VOS animals. Animals were genotyped with more than 40,000 SNPs. Overall, following the implementation of GS in these breeds, we observed a reduction of up to 4 years in generation intervals for sires, fewer calves that later became sires, and, for the French breeds, a broader sire usage. Such changes were likely due to GS enabling the preselection and screening of more young bulls. Additionally, the contributions of the top 10 sires were more evenly distributed after the introduction of GS. Although changes in inbreeding and kinship rates occurred after the introduction of GS, we found no consistent pattern across breeds: pedigree (and genomic runs of homozygosity [ROH]-based) inbreeding rates per generation increased in MRY from -0.67 before GS to 0.51 after GS (from -1.12 to 0.93) and TAR from 0.35 to 0.93 (from 0.68 to 0.86), but decreased in NRC from 0.26 to 0.05 (from 0.10 to 0.06), Abondance from 1.19 to 0.99 (from 2.39 to 0.58), and Vosgienne from 0.53 to 0.23 (from 0.88 to 0.19). Moreover, analysis of genomic ROH-based inbreeding by length class showed that after the implementation of GS, the largest changes in inbreeding level and inbreeding rates per generation occurred for shorter ROH segments. Our study suggests that changes and increases in inbreeding rates may occur after the introduction of GS, although they may not be directly due to the introduction of GS per se, but rather due to population management strategies, such as optimal contribution selection. Our findings emphasize the importance of monitoring changes in both genetic diversity and population demographic structure after implementing GS in local breeds, as well as adjusting breeding strategies when needed to ensure long-term sustainability.

Cet article synthétise une réflexion collective sur les orientations génétiques des bovins laitiers et allaitants de France métropolitaine nécessaires à l’adaptation des animaux aux évolutions des systèmes de production à l’horizon 2040. Il présente les facteurs susceptibles de faire évoluer les productions bovines françaises (démographie des éleveurs, changement climatique, raréfaction des ressources, progrès techniques et évolutions technologiques, réglementations et politiques agricoles, libéralisation du marché, évolution de la demande et organisation des filières). Les systèmes de production bovins devraient néanmoins conserver une diversité importante dans un contexte économique et réglementaire plus contraint, de changement climatique structurant et de raréfaction des ressources non renouvelables. Cet article recommande d’axer la sélection sur la rentabilité économique globale du système plutôt que sur le rendement animal. Pour les bovins laitiers, il s’agira d’augmenter la longévité fonctionnelle, réduire le format des animaux et augmenter la persistance laitière. Pour les bovins à viande, il est souhaitable d’augmenter la précocité sexuelle et la précocité de développement, améliorer la capacité des animaux à valoriser l’herbe et les fourrages, l’aptitude laitière, l’autonomie des animaux et, pour les systèmes d’engraissement herbager, réduire le format. Enfin, pour les bovins laitiers et à viande, il est suggéré d’augmenter la capacité d’ingestion de fourrages grossiers. En regard du changement climatique, il est indispensable de prendre en compte les problèmes sanitaires associés (en particulier émergents), améliorer la tolérance au stress thermique, et réduire les émissions de méthane. La dernière partie de cet article discute des besoins des programmes de sélection induits par ces évolutions.

Gene expression is a dynamic phenotype influenced by tissue-specific regulatory mechanisms, which can modulate expression directly or indirectly through cis or trans factors. Identifying genetic variants in these regulatory regions can improve both expression quantitative trait locus (eQTL) mapping and gene expression prediction. Whole genome sequences offer the possibility for enhanced eQTL mapping accuracy, but detecting causal variants remains challenging. Here, we evaluate the potential added-value of integrating tissue-specific epigenetic annotations, such as chromatin accessibility and methylation status, into within-breed genomic predictions of expression for three pig breeds. Functional annotations from early developmental stages improved eQTL mapping interpretability as shown by the enrichment of trait-relevant QTLs. However, despite the use of functional annotations, predictions across breeds remain challenging due to differences in genetic architectures. Our work contributes to the understanding of gene expression regulation in livestock and highlights the value of functional annotations, despite continued challenges for predictions across breeds.

Genomic prediction (GP) aims to predict the breeding values of multiple complex traits, usually assumed to be multivariate normally distributed by the largely used statistical methods, thus imposing linear genetic relationships between traits. Although these methods are valuable for GP, they do not account for potential nonlinear genetic relationships between traits in scenarios. For individual traits, this oversight may minimally affect prediction accuracy, but it can limit genetic progress when selection involves multiple traits. Deep learning (DL) offers a promising alternative for capturing nonlinear genetic relationships due to its ability to identify complex patterns without prior assumptions about the data structure. We proposed a novel hybrid model that that combines both DL and GBLUP (DLGBLUP), which uses the output of the traditional GBLUP, and enhances its predicted genetic values (PGV) by accounting for nonlinear genetic relationships between traits using DL. We simulated data with linear and nonlinear genetic relationships between traits in order to verify whether DLGBLUP was able to identify nonlinearity when present and avoid inducing it when absent. We found that DLGBLUP consistently provided more accurate PGV for traits simulated with strong nonlinear genetic relationships, accurately identifying these relationships. Over 7 generations of selection, a greater genetic progress was achieved with PGV that accounted for nonlinear relationships (DLGBLUP), compared with GBLUP. When applied to a real dataset from the French Holstein dairy cattle population, DLGBLUP detected nonlinear genetic relationships between pairs of traits, such as conception rate and protein content, and SCC and fat yield, although, no significant increase in prediction accuracy was observed. The integration of DL into GP enabled the modeling of nonlinear genetic relationships between traits, a possibility not previously discussed, given the linear nature of GBLUP. The detection of nonlinear genetic relationships between traits in the French Holstein population when using DLGBLUP indicates the presence of such relationships in real breeding data, suggesting that it may be relevant to further explore nonlinear relationships. This possibility of nonlinear genetic relationships between traits offers a different perspective into multitrait evaluations, with potential to further improve selection strategies in commercial livestock breeding programs. This is particularly relevant when integrating new traits into multitrait evaluations or incorporating new subpopulations, which may introduce different forms of nonlinearity. Finally, it is shown that DL can be used as a complement to the statistical methods deployed in routine genetic evaluations, rather than as an alternative, by enhancing their performance.

Background: Breeding programs select for multiple commercial traits, aiming to achieve genetic progress for all. Often, selection is based on a selection index, i.e. a linear combination of traits with weights defined by, among other information, the genetic correlation between traits. These correlations are typically estimated as a static parameter, and assumed equal to all individuals and generations. While research on the consequences of selection to genetic variances (Bulmer effect) is widely available, only a few studies focused on the consequences of selection to genetic correlations. Our study extended the already existing inferences about how selection affects genetic variances, to how multi-trait selection affects genetic correlations. In order to further our understanding of genetic correlations, we also proposed an alternative method to calculate genetic correlations between traits at the individual level, called by us as individualized sire genetic correlation (iSGC), obtained through the estimated breeding values (EBV) from evaluated daughters. Lastly, a case-study was performed on thirty years of data from the French Holstein dairy cattle population, for five traits studied pairwise: milk and protein yield, milking speed, somatic cell score, and cow conception rate. Results: Theory revealed that multi-trait selection leads to an attenuation (decrease) of positive genetic correlations, with potential to revert them to negative values, if initially low. Uncorrelated traits will become negatively correlated, and negative genetic correlations will be either intensified or attenuated (decrease or increase, respectively), depending on selection intensity, weights applied to the selection index, and the initial genetic correlation. Conclusion: Both theory and empirical results on real data confirm that selection does change the genetic correlation between traits in a population under selection. Moreover, empirical trajectories of the iSGC were in better agreement with the theory, than trajectories of populational genetic correlations. The iSGC searches for individual-specific patterns of correlations, and since it is measured on sires through the EBV of their daughters, it also considers the recombination of the genetic background. Along with the fact that trajectories of iSGC were in better agreement with theory, we believe it to be a potentially less biased measure of genetic correlations between traits. ### Competing Interest Statement The authors have declared no competing interest.

Most validation studies of genomic evaluations on candidates (prior to observing phenotypes) present inflation of their predicted breeding values, i.e., regression coefficients of their later observed phenotypes on the early predictions are smaller than one. The aim of this study was to show that this inflation pattern reflects at least partly long-distance associations between markers and quantitative trait loci (QTL) in the reference population and to propose methods to estimate the corresponding “erosion” coefficient. Across-chromosome linkage disequilibrium (LD) is observed in different dairy cattle breeds, being a result from limited effective population size and from relationships within the reference population. Due to this long distance LD, the estimated SNP effects capture non-zero contributions from distant QTLs, some located on other chromosomes than the SNP itself. Therefore, corresponding SNP effects are partly lost in the next generations and we refer to this loss as “erosion”. With the concept of QTL contribution to SNP effects derived from mixed model equations, we show with simulation that this long range LD explains 6–25

Juvenile mortality is a major concern for the dairy cattle industry because of its wide-ranging economic, environmental, and ethical effects. Over the last few years, an increased number of genetic defects associated with juvenile mortality has been identified, creating new opportunities and challenges for their management in selection. The implementation of a routine genomic evaluation may therefore contribute to reducing juvenile mortality. However, no existing method currently includes causal variants and incorporates their additive and dominance effects in a single-step GBLUP (ssGBLUP) model. This study was conducted on 2 French dairy breeds: Holstein (HOL) and Montbéliarde (MON), both of which are affected by known recessive genetic defects (bovine lymphocyte intestinal retention defect and cholesterol deficiency in HOL, and mitochondropathy in MON). In this study, we first estimated genetic parameters for juvenile mortality. Second, we developed ssGBLUP models that included causal variants, either by assigning higher proportions of genetic variance explained a priori by the causal variants or by estimating and incorporating causal variants' additive and dominance effects as covariates. Third, we evaluated the models' performance to select the most accurate one for genomic evaluation. To this end, 2 independent training and validation datasets were constructed. The training dataset included juvenile mortality records for 1,275,746 HOL and 476,361 MON females, of which 84,328 HOL and 29,944 MON were genotyped. Models were validated using mortality records of 689,502 HOL and 149,801 MON daughters of 262 and 254 genotyped bulls, respectively. No difference in model performance was seen at the population level. However, at the individual level, heterozygous bulls for the causal variants were more accurately distinguished from wild type homozygous when using EBVs obtained from models that included causal variants compared with a conventional ssGBLUP. We therefore advocate the routine implementation of a genomic evaluation that accounts for causal variants to help reduce juvenile mortality. A genotyping effort for dead animals is also suggested because it would enhance model performance. Additionally, further research is needed to determine the most effective method for incorporating causal variants into ssGBLUP models and to assess whether their inclusion improves the model's prediction accuracy.

The performance of dairy cows is influenced by the microbial communities hosted within their digestive tract. While the rumen microbiota has long been associated with host phenotypes, the impact of the faecal microbiota remains elusive. In this study, we collected 697 faecal samples from commercial Holstein cows and analysed them with 16S rRNA gene analyses. For each animal, routinely recorded data, i.e., milk yield, fat yield, protein yield, fat content, protein content, and an aggregate production trait (pINEL) based on the French economic dairy index, were available to assess the links between the faecal microbiota and host production. Our findings revealed a strong and significant association between the structure of the bacterial and prokaryote community (β-diversity) and dairy production. In addition, differential abundance analyses identified 48 genera whose abundances were significantly associated with pINEL, milk, fat and protein yield. Among these genera, the increased abundance of Bifidobacterium, and particularly an amplicon sequence variant with a 16S rRNA V3-V4 gene region identical to B. globosum and B. pseudolongum, was found to be the most important for high-yielding animals. Bifidobacterium seemed to be a potential key member of the bovine faecal microbiota that should be further investigated. Conversely, the p-1088-a5 gut group genus was found more abundant in low-productive cows. In conclusion, this study demonstrates significant associations between the faecal microbiota and the performance of dairy cows at the whole lactation scale. A better understanding of the physiology of the gut microbiota could help to improve dairy cow production.

Due to their potential impact on the host's phenotype, organ-specific microbiotas are receiving increasing attention in several animal species, including cattle. Specifically, the vaginal microbiota of ruminants is attracting growing interest, due to its predicted critical role on cows' reproductive functions in livestock contexts. Notably, fertility disorders represent a leading cause for culling, and additional research would help to fill relevant knowledge gaps. In the present study, we aimed to characterize the vaginal microbiota of a large cohort of 1171 female dairy cattle from 19 commercial herds in Northern France. Vaginal samples were collected using a swab and the composition of the microbiota was determined through 16S rRNA sequencing targeting the V3-V4 hypervariable regions. Initial analyses allowed us to define the core bacterial vaginal microbiota, comprising all the taxa observed in more than 90% of the animals. Consequently, four phyla, 16 families, 14 genera and a single amplicon sequence variant (ASV) met the criteria, suggesting a high diversity of bacterial vaginal microbiota within the studied population. This variability was partially attributed to various environmental factors such as the herd, sampling season, parity, and lactation stage. Next, we identified numerous significant associations between the diversity and composition of the vaginal microbiota and several traits related to host's production and reproduction performance, as well as reproductive tract health. Specifically, 169 genera were associated with at least one trait, with 69% of them significantly associated with multiple traits. Among these, the abundances of Negativibacillus and Ruminobacter were positively correlated with the cows' performances (i.e., longevity, production performances). Other genera showed mixed relationships with the phenotypes, such as Leptotrichia being overabundant in cows with improved fertility records and reproductive tract health, but also in cows with lower production levels. Overall, the numerous associations underscored the complex interactions between the vaginal microbiota and its host. Given the large number of samples collected from commercial farms and the diversity of the phenotypes considered, this study marks an initial step towards a better understanding of the intimate relationship between the vaginal microbiota and the dairy cow's phenotypes.

Background Genomic prediction aims to predict the breeding values of multiple complex traits assumed to be normally distributed, thus imposing linear genetic correlations between traits. However, these statistical methods are unable to model nonlinear genetic relationships between traits, if existent, potentially leading to a decrease in prediction accuracy. Deep learning (DL) is a promising methodology for predicting multiple complex traits, in scenarios where nonlinear genetic relationships are present, due to its capacity to capture complex and nonlinear patterns in large data. We proposed a novel pure DL model, designed to obtain predicted genetic values (PGV) while accounting for nonlinear genetic relationships between traits, and extended this model to a hybrid DLGBLUP model which uses the output of the traditional GBLUP, and enhances its PGV by using DL. Using simulated data, we compared the accuracy of the PGV obtained with the proposed pure DL model, the hybrid DLGBLUP model, and the traditional GBLUP model – the latter being our baseline reference. Results We found that both DL and DLGBLUP models either outperformed GBLUP, or presented equally accurate PGV, with a particular greater accuracy for traits presenting a strongly characterized nonlinear genetic relationship. DLGBLUP presented the highest prediction accuracy and smallest mean squared error of the PGV for all traits. Additionally, we evolved a base population over seven generations and compared the genetic progress when selecting individuals based on the additive PGV obtained by either DL, DLGBLUP or GBLUP. For all traits with a nonlinear genetic relationship, after the fourth generation, the observed genetic gain when selection was based on the additive PGV from GBLUP was always inferior to the observed when selection was based on either DL or DLGBLUP. Conclusions The integration of DL into genomic prediction has potential to bring significant advancements in the field. By identifying nonlinear genetic relationships, our DL and DLGBLUP models improved prediction accuracy. It offers an insight to genetic relationship and its evolution over generations, with potential to improve selection strategies in commercial livestock breeding programs. Moreover, DLGBLUP shows that DL can be used as a complement to statistical methods, by enhancing their performance. ### Competing Interest Statement The authors have declared no competing interest.

The fecal microbiota of ruminants constitutes a diversified community that has been phenotypically associated with a variety of host phenotypes, such as production and health. To gain a better understanding of the complex and interconnected factors that drive the fecal bacterial community, we have aimed to estimate the genetic parameters of the diversity and composition of the fecal microbiota, including heritabilities, genetic correlations among taxa, and genetic correlations between fecal microbiota features and host phenotypes. To achieve this, we analyzed a large population of 1,875 Holstein cows originating from 144 French commercial herds and routinely recorded for production, somatic cell score, and fertility traits. Fecal samples were collected from the animals and subjected to 16S rRNA gene sequencing, with reads classified into Amplicon Sequence Variants (ASVs). The estimated α- and β-diversity indices (i.e., Observed Richness, Shannon index, Bray-Curtis and Jaccard dissimilarity matrices) and the abundances of ASVs, genera, families and phyla, normalized by centered-log ratio (CLR), were considered as phenotypes. Genetic parameters were calculated using either univariate or bivariate animal models. Heritabilities estimates, ranging from 0.08 to 0.31 for taxa abundances and β-diversity indices, highlight the influence of the host genetics on the composition of the fecal microbiota. Furthermore, genetic correlations estimated within the microbial community and between microbiota features and host traits reveal the complex networks linking all components of the fecal microbiota together and to their host, thus strengthening the holobiont concept. By estimating the heritabilities of microbiota-associated phenotypes, our study quantifies the impact of the host genetics on the fecal microbiota composition. In addition, genetic correlations between taxonomic groups and between taxa abundances and host performance suggest potential applications for selective breeding to improve host traits or promote a healthier microbiota.

BACKGROUND:Combining the results of within-population genome-wide association studies (GWAS) based on whole-genome sequences into a single meta-analysis (MA) is an accurate and powerful method for identifying variants associated with complex traits. As part of the H2020 BovReg project, we performed sequence-level MA for beef production traits. Five partners from France, Switzerland, Germany, and Canada contributed summary statistics from sequence-based GWAS conducted with 54,782 animals from 15 purebred or crossbred populations. We combined the summary statistics for four growth, nine morphology, and 15 carcass traits into 16 MA, using both fixed effects and z-score methods.RESULTS:The fixed-effects method was generally more informative to provide indication on potentially causal variants, although we combined substantially different traits in each MA. In comparison with within-population GWAS, this approach highlighted (i) a larger number of quantitative trait loci (QTL), (ii) QTL more frequently located in genomic regions known for their effects on growth and meat/carcass traits, (iii) a smaller number of genomic variants within the QTL, and (iv) candidate variants that were more frequently located in genes. MA pinpointed variants in genes, including MSTN, LCORL, and PLAG1 that have been previously associated with morphology and carcass traits. We also identified dozens of other variants located in genes associated with growth and carcass traits, or with a function that may be related to meat production (e.g., HS6ST1, HERC2, WDR75, COL3A1, SLIT2, MED28, and ANKAR). Some of these variants overlapped with expression or splicing QTL reported in the cattle Genotype-Tissue Expression atlas (CattleGTEx) and could therefore regulate gene expression.CONCLUSIONS:By identifying candidate genes and potential causal variants associated with beef production traits in cattle, MA demonstrates great potential for investigating the biological mechanisms underlying these traits. As a complement to within-population GWAS, this approach can provide deeper insights into the genetic architecture of complex traits in beef cattle.

Background Because of its potential influence on the host’s phenotype, increasing attention is paid to organ-specific microbiota in several animal species, including cattle. However, ecosystems other than those related to the digestive tract remain largely understudied. In particular, little is known about the vaginal microbiota of ruminants despite the importance of the reproductive functions of cows in a livestock context, where fertility disorders represent one of the primary reasons for culling.Results In the present study, we aimed at better characterizing the vaginal microbiota of dairy cows through 16S rRNA sequencing, using a large cohort of Holstein cows from Northern France. Our results allowed to define a core microbiota of the dairy cows’ vagina, and highlighted that 90% of the sequences belonged to the Firmicutes, the Proteobacteria, and the Bacteroidetes phyla. The core microbiota was composed of four phyla, 16 families, 14 genera and only one amplicon sequence variant (ASV), supporting the idea of the high diversity of vaginal microbiota within the studied population. This variability was partly explained by various environmental factors such as the herd, the sampling season, the lactation rank and the lactation stage. In addition, we investigated potential associations between the diversity and the composition of the vaginal microbiota and several health-, performance-, and fertility-related phenotypes. Our analyses highlighted significant associations between the α and β- diversities and several traits including the first insemination outcome, the productive longevity, and the culling. Besides, relevant phenotypes were correlated with the abundance of several genera, some of which, such as Leptotrichia , Streptobacillus , Methylobacterium-Methylorubrum , or Negativibacillus , were linked to multiple traits.Conclusion Considering the large number of samples, which were collected in commercial farms, and the diversity of the phenotypes considered, this study represents a first step towards a better understanding of the close relationship between the vaginal and the dairy cow’s phenotypes.### Competing Interest StatementLB, SD, GE, SMa, SMe and CA were employed by GD Biotech / Genes Diffusion company. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.* AI : Artificial insemination ANCOM-BC : Analysis of compositions of microbiomes with bias correction ANOVA : Analysis of variance ASV : Amplicon sequence variant C-AIf : Calving to fertilizing insemination interval CI : Calving interval DADA2 : Divisive amplicon denoising algorithm 2 FIS : First insemination success OTU : Operation taxonomic variant PERMANOVA : Permutational multivariate analysis of variance

In recent years, crossbreeding strategy has grown in dairy cattle farms at an international level. Breeders are interested in keeping crossbred cows in their herd both to combine the strengths of the pure breeds, compensate for their weaknesses and benefit from heterosis. However genetic tools are still lacking to manage these crossbred animals. In this study, we evaluate the performances of a genomic evaluation adapted for rotational crossbreeding schemes with real data. This genomic evaluation was applied to a population that includes pure-breed animals from Holstein, Montbéliarde, and Red Danish breeds, as well as crossbreds between these three breeds. The genomic evaluation approach was based on the estimation of breed specific SNP specific according to the Breed of Origin of the Alleles.

Background: It is now widespread in livestock and plant breeding to use genotyping data to predict phenotypes with genomic prediction models. In parallel, genomic annotations related to a variety of traits are increasing in number and granularity, providing valuable insight into potentially important positions in the genome. The BayesRC model integrates this prior biological information by factorizing the genome according to disjoint annotation categories, in some cases enabling improved prediction of heritable traits. However, BayesRC is not adapted to cases where markers may have multiple annotations. Results: We propose two novel Bayesian approaches to account for multi-annotated markers through a cumulative (BayesRC+) or preferential (BayesRCπ) model of the contribution of multiple annotation categories. We illustrate their performance on simulated data with various genetic architectures and types of annotations. We also explore their use on data from a backcross population of growing pigs in conjunction with annotations constructed using the PigQTLdb. In both simulated and real data, we observed a modest improvement in prediction quality with our models when used with informative annotations. In addition, our results show that BayesRC+ successfully prioritizes multi-annotated markers according to their posterior variance, while BayesRCπ provides a useful interpretation of informative annotations for multi-annotated markers. Finally, we explore several strategies for constructing annotations from a public database, highlighting the importance of careful consideration of this step. Conclusion: When used with annotations that are relevant to the trait under study, BayesRCπ and BayesRC+ allow for improved prediction and prioritization of multi-annotated markers, and can provide useful biological insight into the genetic architecture of traits.