Current status of genomic prediction using Multi-omics data in livestock

An, Na Rae; Lee, Seung Su; Park, Jong Eun; Chai, Han Ha; Cho, Yong Min; Lim, Dajeong

Abstract

It became possible to perform genomic predictions using single nucleotide polymorphism (SNP) with advancements in genomics technology, not only in human but in livestock as well. There are strong interests in improving economical traits in livestock through identifying causative mutation, genes or predicting genomic breeding values. We present the current status of genome prediction studies for phenotype estimation of economic traits in livestock from various perspectives based on the genomic area. First, we introduce theoretical background of genomic prediction methods and newest development on SNP information. Thanks to develop sequencing technology, multi-omics data can be used to predict phenotypes associated with the economic traits. In particular, many studies show that genomic prediction accuracy of genomic partitioning data based on the biological information is higher than that of commercial SNP chip. Therefore, multi-omics data can be useful for genomic prediction studies. It is also important that researchers should consider factors affecting genomic prediction accuracy such as heritability, Quantitative Trait Loci (QTL) and marker density, size and structure of reference population. We also introduce genomic prediction studies based on the integration of multi-omics data that shows improvement of prediction accuracy than typical Genomic Best Linear Unbiased Prediction (GBLUP) models. We concluded that genomic prediction studies can be expanded to apply social issues, new phenotypes, or precision agriculture such as diseases, climate change, and metabolism including economic traits with multi-omics data using high-throughput technologies.