Multi-omics analysis

Integrating biological annotations to improve genomic prediction in pigs

---

Abstract

Genomic prediction has been a pivotal method in animal breeding for more than a decade. As recent studies increasingly highlight the role of biological annotations in trait-specific regulation, there is growing interest in developing approaches that use these annotations to improve prediction accuracy. Incorporating biological annotations allows us to relax the assumption that every locus is equally likely to affect the trait, that is, that all loci share the same prior distribution. Because this assumption is biologically unrealistic, biological annotations provide information that is independent of linkage disequilibrium and offer insight into the likely locations and effects of causal variants. To address this, we have developed and implemented annotation dependent Bayesian models in which the prior for each SNP is determined by the annotations assigned to that SNP. We are proposing this model to evaluate whether it can improve the prediction of pig traits compared with conventional genomic prediction models. To evaluate the model, we used a population of about 89,000 pigs genotyped for approximately 30 million SNPs. After quality control, 2.5 million SNPs remained across nine commercially important traits. We considered four annotations in the analysis: Intron, Coding Sequence(CDS), Untranslated Region(UTR), and all remaining regions. Compared with a conventional model without annotations, our annotation-dependent Bayesian model showed trait-specific gains in prediction accuracy for three out of nine traits. For Number of Stillborn, prediction accuracy increased from 0.273 to 0.283 (a 3.6% improvement). For Litter Weight at Weaning, accuracy increased from 0.167 to 0.174 (a 4.2% improvement). For Return to Estrus seven days post-weaning, accuracy increased from 0.085 to 0.097 (a 14.1% improvement). These preliminary findings suggest that biological annotations may contribute to genomic prediction in pigs. This impact will be further evaluated using more comprehensive and precise annotation information.

Keywords: 2026

How to Cite:

Liang, D., Qu, J., Roddur, Q., Valente, B., Chen, C., Carrara, E., Lourenco, D., Holl, J. & Cheng, H., (2026) “Integrating biological annotations to improve genomic prediction in pigs”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2287221. doi: https://doi.org/10.31274/wcgalp.24238

Rights: 1