New genomic regions associated with litter size and its variation in pigs

One of the main breeding goals in pig breeding is increasing total number born in a litter (TNB). Variation in this trait is large and increasing TNB in a population could exceed the physiological capacity of sows to provide for a litter pre- and post-farrowing. Thus there is a desire to breed for increased mean TNB while at the same time reducing its variability. Here we study the variation of TNB in a Large White pig population by applying Double Hierarchical GLM (DHGLM) and a genome-wide association study (GWAS). The residual variance of TNB (varTNB) and its variance components were estimated with DHGLM in ASReml. For this step, 263,088 observations on TNB were available. Estimated breeding values (EBV) obtained with DHGLM were used to calculate deregressed EBV (dEBV) for 2,351 sows and boars genotyped with 64k chip. The GWAS was performed with a Bayesian Variable Selection method in Bayz. SNP’s were considered significant if their Bayes Factor was above 30. Genetic coefficient of variation of the standard deviation for varTNB was estimated as 0.09, indicating good opportunities for improvement of uniformity by selection. Genetic correlation between additive genetic effects onTNB and on its variation was 0.5. This indicates that the selection for TNB increases the variation of TNB. Ten SNPs were detected for TNB and nine SNPs for varTNB. The most significant SNP explained 0.4% of genetic variance in TNB (SSC11) and 0.5% in varTNB (SSC7). Possible candidate genes for varTNB on SSC7 are: heat shock protein (HSPCB alpha), vascular endothelial growth factor (VEGFA), and CUL9 protein regulating p53 function. This is the first study reporting SNPs and candidate genes associated with varTNB in pigs.