Genome-wide association study in Brown Swiss for udder traits based on sequence data

Identification of QTL, especially of causative variants within QTL is still challenging. Higher SNP densities aid the identification and fine-mapping of QTL. Based on imputed sequence data we performed GWAS for udder traits in Brown Swiss cattle. The GWAS was performed using a mixed-model approach with deregressed breeding values as phenotypes. The traits investigated included: udder depth (UD), fore udder attachment (FUA), rear udder width (RUW), rear udder height (RUH), fore udder length (FUL), and central ligament (CL). We found significant associations on BTA 3 (UD, FUA), BTA 5 (UD), BTA 17 (FUL, RUW, CL) and BTA 20 (FUA). A single gene was located in the significantly associated regions on BTA 5 (ABCC9) and BTA 20 (HCN1). The region on BTA 17 spans almost 3 Mb and includes 74 genes (maximal region for all the traits combined) and the region on BTA 3 includes 91 genes across 3 Mb. We also looked for associated missense variants in these intervals. Neither for ABCC9 nor for HCN1 we could identify such a variant. On BTA 17 we identified 2 missense variants that were significantly associated with CL. On BTA 3, 11 missense variants were significantly associated with UD and/or FUA. The advantages of using imputed sequence data compared to SNP chip genotypes are mainly through the inclusion of potential causative variants.