Weighted single-step GBLUP improves accuracy of genomic predictions for traits controlled by major gene or QTL in French dairy goats

First studies have investigated the feasibility of genomic evaluation in French dairy goat (Carillier et al., 2014).They have shown the interest of ssGBLUP method (Legarra et al., 2009), using performances of all females, genotypes and pedigree. This method was then implemented in French dairy goat in 2017. It assumes that traits are under a complete polygenic determinism. However, in the two main French dairy goat breeds (Alpine and Saanen), QTL and major genes have been identified for traits, such as αs1 casein gene (chromosome 6) or DGAT1 (chromosome 14) (Martin et al., 2017). Taking into account this information in a model to estimate genomic breeding values could increase accuracy of genomic predictions of candidates. Weighted ssGBLUP approaches developed by Wang et al. (2012) were proposed to give more weights to SNP strongly associated to a trait in the construction of the relationship matrix. In this study, ssGBLUP, WssGBLUP and three variants of WssGBLUP were compared to evaluate the benefit of taking into account QTL and/or major gene in the French dairy goat genetic evaluation for five milk productions, five udder types and one functional traits.