116 Contribution of X-chromosome to the genetic variance of litter size and stillbirth in pigs

Abstract

Abstract Genes with important roles in reproduction traits have been described in pig X-chromosome (SSCX), which may contribute considerably to genetic variance and could impact the accuracy of genomic prediction. The phenotypic data set includes information on 82,580 litter size records from 34,346 sows. Additionally, there were 108,006 in the pedigree, and genotypes for 41,492 SNPs, of which 1,435 in SSCX, were available for 30,000 animals. Three different approaches were used to estimate variance components for the total number born and number of stillborn. In the first approach, only SNPs in the autosomes were used and the model included the additive genetic (u), permanent environment (p), and residual (e) random effects. In the second approach, the same model was adopted; however, the data included SNPs in the autosomes and in the X chromosome. For both models, it was assumed that u~N(0,H⊗Vu), e~N(0,I⊗Ve) and p~N(0,I⊗Vp), wherein H is a relationship matrix that combines pedigree and genomic information; and, Vu, Vp, and Ve are the (co)variance matrices for the random additive genetic, permanent environment, and residual effects, respectively. In the third approach, the random additive genetic effects associated with the X-chromosome (s) were included in the model described before. Then, it was assumed that s~N(0,Hx⊗Vs), where Hx is the genomic relationship matrix for the X-chromosome and Vs is the additive genetic variance associated with the SSCX. All analyses were carried out using GIBBSF90+. Table 1 shows variance components for the three approaches. For both traits, additive genetic variance and heritabilities were similar across the approaches. The additive genetic variance for the SSCX was small, accounting for approximately two percent of the total additive genetic variance. Including the X chromosome as a separate effect and using Hx as a covariance matrix captured only a small part of the additive genetic variance. The additive variance across approaches did not differ. For total number born and number of stillborn, using the SSCX SNPs to build Hx did not provide extra benefit compared with fitting those SNP in the regular H matrix.