La base documentaire de l'IFIP

La base documentaire de l'IFIP : des centaines de documents à télécharger ou bien à commander.

Résultats 1 à 3 de 3 résultats
Rechercher une documentation
Publication Annéetrier par ordre croissant

Importance du phénotypage pour maintenir la précision des prédictions génomiques des caractères mesurés en station

Consulter le resumé

Céline Carillier-Jacquin (Inrae) et al., 52es Journées de la Recherche Porcine (FRA), 4 et 5 février 2020, poster

Poster.

Document réservé Espace Pro, veuillez vous identifier
2020

Importance du phénotypage pour maintenir la précision des prédictions génomiques des caractères mesurés en station

Consulter le resumé

Céline Carillier-Jacquin (Inrae) et al., 52e Journée de la Recherche Porcine (FRA), 4 et 5 février 2020, Paris, p. 43-44, poster

Poster.

L’évaluation génomique dans les lignées porcines maternelles françaises a été mise en place en 2016 et a permis d’augmenter le progrès génétique, notamment pour les caractères de reproduction. Cependant, des problèmes calculatoires sont apparus pour l’évaluation génomique de certains caractères de production mesurés en station, dont la capacité de phénotypage est limitée. Cela concerne particulièrement les candidats génotypés des élevages de sélection qui ne disposent pas de phénotypes en station et peu d’apparentés phénotypés. La structure de données pour ce type de caractère (i.e. porcs génotypés (candidats) non phénotypés et porcs phénotypés (animaux station) non génotypés) pourrait être responsable des problèmes de convergence observés pour la prédiction des valeurs génomiques. Dans cette étude, nous avons simulé, à partir de phénotypes réels mesurés sur l’ensemble des candidats à la sélection, différents scénarios pour mimer la situation de phénotypage partiel rencontrée dans le cas des caractères mesurés en station et évaluer l’impact d’une augmentation du nombre de porcs phénotypés sur la précision des valeurs génomiques obtenues.

ENG

Poster.

Importance of phenotyping to maintain the accuracy of genomic predictions of traits measured in a test station

Genomic evaluation of French maternal lines, set up in 2016, has helped increase genetic progress, especially for reproductive traits. However, computational problems have emerged for genomic evaluation of certain production traits for which phenotyping capacity is limited. This particularly concerns genotyped candidates on breeding farms that have no phenotypes and only a few phenotyped relatives. This data structure seems to pose convergence problems for predicting genomic breeding values. To check this hypothesis, we simulated such a situation, based on a set of actual phenotype data measured for all farm candidates and genotypes. The simulation consisted of deleting phenotypes of the animals measured on-farm in order to reproduce the data structure encountered for the traitsrecorded at the FGPorc/INRAE test station in Le Rheu. Phenotypes were then progressively added in different scenarios to identify whether prediction accuracy improved and to estimate the number of phenotypes required. The simulations showed that the unbalanced structure between genotypes and phenotypes was responsible for the computational problems that led to low accuracy of genomic predictions. Phenotyping 12% of all pigs phenotyped at 100 kg each year made it possible to solve the computational problems observed and to recover 61% of the maximum expected accuracy. In conclusion, these results highlight the importance of collecting large-scale phenotypes in the context of genomic selection schemes. Further studies will be conducted to study the impact of genotyping animals measured at the station.

Document réservé Espace Pro, veuillez vous identifier
2020

Redlosses: reducing food losses by microbial spoilage prediction

Consulter le resumé

M. Zagorec (INRA) et al., EFFOST, 6-8 novembre 2018, Nantes

Food spoilage leads to significant wastes and losses, and is an important economic issue in food industry. In the case of meat, a large part of spoilage is the consequence of bacterial growth and subsequent metabolic activities causing organoleptic spoilage of the final product (defects in texture, color, odor, or aspect), leading finally to products that are lost because they do not fit the quality standards. In addition, meat production chain requires energy, water and cost consuming operations (i.e. animal breeding, slaughtering, and transformation and storage which are usually performed at low temperature). Therefore meat product spoilage that appears at the end of the process or during shelf life affects the whole production chain performances as well as the sustainability label of the meat sector. The objective of the project is to reduce food losses by predicting, early in the production process, the onset of bacterial spoilage during storage in order to propose decision-support tools for directing process. Pork and poultry meat, the two main meats consumed in France will be studied. The economic impact of losses of these products will be assessed. Dynamics of bacterial communities will be monitored during processing steps (from primary cuts to end products at use-by-date and beyond) and various descriptors of spoilage will be measured. The natural variability between batches and that associated with production processes will be considered. Data will be used to identify accurate spoilage markers and to compute innovative mathematical models for predicting spoilage occurrence as a function of the initial composition of the microbiota (diversity and abundance) and some abiotic factors (lactate concentration, nature of packaging atmosphere). The models will be validated on meat products, including the economic aspect in order to propose decision-support tools for the food producers. The project and first results will be presented.

Document réservé Espace Pro, veuillez vous identifier
2018