Using transcriptomic data to improve the prediction of immunity traits in pigs
Considering health-related traits among breeding selection criteria has been proposed as a way to improve pig robustness. This study investigated the potential of whole blood RNA-sequencing data for predicting immunity-related traits, stress indicators and carcass weight, using data from 255 pigs be...
| Autores: | , , , , |
|---|---|
| Tipo de recurso: | artículo |
| Fecha de publicación: | 2025 |
| País: | España |
| Institución: | Varias* (Consorci de Biblioteques Universitáries de Catalunya, Centre de Serveis Científics i Acadèmics de Catalunya) |
| Repositorio: | Recercat. Dipósit de la Recerca de Catalunya |
| OAI Identifier: | oai:recercat.cat:20.500.12327/5015 |
| Acceso en línea: | https://hdl.handle.net/20.500.12327/5015 https://doi.org/10.1016/j.animal.2025.101742 |
| Access Level: | acceso abierto |
| Palabra clave: | 577 |
| Sumario: | Considering health-related traits among breeding selection criteria has been proposed as a way to improve pig robustness. This study investigated the potential of whole blood RNA-sequencing data for predicting immunity-related traits, stress indicators and carcass weight, using data from 255 pigs belonging to a commercial Duroc population. The prediction performance of mixed models fitting either genomic (G), transcriptomic (T) or both effects as independent (GT) was evaluated and compared. Three additional models addressing the redundant information between G and T were also evaluated: the GTC model that subtracts the genetic effect from the transcriptome, the GTCi model that makes this correction based on the estimated heritability of T effects, and a multiomic model that weights G and T effects in a multiomics relationship matrix. The models including gene expression information captured a higher proportion of variance than the genomic model for all studied traits but carcass weight. Adding transcriptomic effects improved both model fit and phenotypic prediction of all immunity traits, particularly those with a high transcriptomic contribution such as the abundance of T helper and γδ T cells, the haptoglobin concentration and the leukocyte counts. Considering the interaction between genomic and transcriptomic effects led to greater prediction accuracies, with the GTCi model performing the best. Our work demonstrates the value of considering gene expression data to predict immunity traits as well as the importance of adequately modelling the interaction between genomic and transcriptomic effects. |
|---|