Genome-wide association analysis in Italian Simmental cows for lactation curve traits using a low density (7K) SNP panel

Abstract Text: In genomic selection (GS) programs of dairy cattle breeds, bulls are currently genotyped using medium or high density SNP platforms. Genome wide association studies (GWAS) are then carried out on variables (as polygenic breeding values or deregressed proofs) that are derived from phenotypes recorded on females. It could be therefore of interest to perform GWAS directly on cows, that now are becoming to be genotyped in GS programs often by using low density SNP panels. In this study GWAS was performed on 1,211 lactations of 337 Italian Simmental cattle, genotyped with the 7K Illumina bead-chip. Edit on markers was performed on call rate (>.99) and minor allele frequency (>.01). After edits, 6,891 markers were retained for the analysis. Phenotypes were scores of the first two principal components obtained by performing principal component analysis on the test day records (seven for each lactation) for milk yield, fat and protein percentages, and somatic cell scores (SCS). The first component describes the average level (LEVEL) of the lactation curve, i.e. the average level of milk yield and percentages along the whole lactation, whereas the second is related to the shape of the lactation curve (SHAPE). Data were analysed with a mixed linear model that included fixed effects of calving month, calving year, parity, SNP genotype (coded as 0,1,2), and  random effects of herd, animal additive,  permanent environment. Multiple testing was accounted  by performing the Bonferroni correction (uncorrected P x number of tests). Two markers were associated to LEVEL for milk yield according to the Bonferroni corrected statistical significance (P<0.05). One was located on BTAs 6, close to the casein cluster. Nine significant SNPs were highlighted for LEVEL of fat percentage. Most  important (P<0.001) were on BTA23, close to the desmoplakin gene which is involved in the turnover of epithelial cells, and on BTA7 close to the Ring Finger protein 145. Moreover, two significant SNP were located on BTA14, close to the CYP11B1 gene, previously reported to be associated to dairy traits in Holstein. Six markers were associated to protein percentage. The most significant was on BTA12, close to the HSTRA locus, reported to be associated to type traits in Chinese Holstein. Finally, one marker was found to affect SCS on BTA22, close to the LANCL2 locus.  No markers were found to be associated to the SHAPE component for any of the four considered traits 

Keywords: GWAS, LD panel, Italian Simmental Cattle