RNA Sequencing of 150 Lambs Connects Meat Phenotypes with Gene Expression

Gene expression analysis can aid in prioritising regions or classes of variants for genomic prediction and help increase the understanding of quantitative traits. Liver and muscle tissue were collected at slaughter for 150 lambs, with up to 70 meat traits recorded per lamb. Their dams were managed to high, medium, and low body condition scores (BCS) during mid-to-late pregnancy with the lambs finished on three different diets. Differential expression of genes (DEG) investigated contrasting tissue, BCS, lamb diets, other treatment differences, as well as high and low lamb carcass eye muscle width (CEMW).  Covariance matrices of depth normalized read counts per exon were fitted in a mixed model to estimate the proportion of the phenotypic variance explained by co-expression.  Expression QTL (eQTL) analyses were performed on the Ovine high density SNP chip and on the newly discovered SNP using RNA sequence.  A large number of DEG were identified between tissues, but only the low versus high BCS comparison resulted in any DEG for treatments. DEG was also found for lambs with high and low CEMW. A strong trend toward down regulation was observed in all tests, except for BCS, where all DEG were overexpressed in lambs from ewes with higher condition score during mid-late pregnancy.  Co-expression accounted for up to 79% of the phenotypic variance but was very variable across traits.  Many eQTL passed multiple testing and are now being compared to results from an independent meat trait genome-wide association study of 10,000 lambs.