Using transcriptomic data to improve the prediction of immunity traits in pigs

Area of Science:

• Animal Genetics and Genomics
• Immunology
• Bioinformatics

Background:

• Improving pig robustness through breeding selection is crucial for animal health.
• Health-related traits are increasingly considered in breeding programs.
• Whole blood RNA sequencing offers potential for predicting complex traits.

Purpose of the Study:

• To evaluate the predictive performance of transcriptomic data for immunity-related traits, stress indicators, and carcass weight in pigs.
• To compare genomic (G), transcriptomic (T), and combined (GT) models for trait prediction.
• To assess advanced multiomic models accounting for G-T interactions.

Main Methods:

• Utilized whole blood RNA sequencing data from 255 commercial Duroc pigs.
• Employed mixed models including genomic, transcriptomic, and combined effects (G, T, GT).
• Evaluated models like GTC, GTCi, and a multiomics relationship matrix model to handle G-T redundancy and interactions.

Main Results:

• Models incorporating gene expression data explained more variance than genomic models for immunity and stress traits, excluding carcass weight.
• Transcriptomic effects significantly improved model fit and prediction accuracy for immunity traits, especially T helper cell abundance, γδ T cell abundance, haptoglobin, and leukocyte counts.
• Models accounting for genomic-transcriptomic interactions, particularly the GTCi model, achieved the highest prediction accuracies.

Conclusions:

• Whole blood gene expression data is valuable for predicting pig immunity traits.
• Accurate modeling of interactions between genomic and transcriptomic effects is essential for maximizing prediction accuracy in multiomic studies.
• This approach enhances the potential for improving pig robustness and health through advanced breeding strategies.