Epsilon Archive for Student Projects

Abstract

Genomic selection has been around for more than 10 years in the pig industry. Despite
the popularity, accurately predicting complex with genome-wide markers is still
challenging. The current limitation in prediction accuracy is, at least partially, due to the
use of markers that may not directly influence the traits of interest. Some studies suggest
that whole-genome sequencing can improve genomic predictions within and across
breeds, but empirical results are inconclusive. Another proposed solution to the problem
is the use of functional genomics in genomic selection. Functional genomics studies genes
and their interactions using techniques like expression quantitative trait loci (eQTL) and
chromatin immunoprecipitation sequencing. Among these techniques, eQTL studies
identify genomic regions influencing gene expression levels, which in turn affect
conventional phenotypes. Given this complex connection among genomic variants,
regulation of gene expression and phenotype formation, there is, therefore, a growing
recognition of the potential benefit of integrating functional genomic information,
particularly expression quantitative trait loci, into genomic selection models to enhance
prediction accuracy. This study was performed in an attempt to find out under what
circumstances eQTL markers will improve selection accuracy in genomic selection for
pig breeding.
The project considered a simulation of a breeding programme selecting a single
polygenic trait affected by 50 genes. Using the R package “AlphaSimR” the pig genome
and base population were obtained from a coalescent simulation and used to simulate
breeding activities. An eQTL mapping study was performed using a simulated 100k chip.
Genomic selection was performed under three scenarios. Two of the scenarios used
variants from the eQTL study as markers, and the third used simulated SNP markers
represented as 1) eQTL only, 2) eQTL plus SNP, and 3) SNP only. Results did not
indicate any big impact of eQTL markers on genomic selection, despite a successful
eQTL mapping. The project highlighted some limitations indicating that further
investigation of eQTLs in genomic selection is necessary.