GWAS for genotype by lactation stage interaction for milk production

The genetic background of milk production traits changes during lactation. However, most genome-wide association study (GWAS) for milk production traits assume that genetic effects are constant during lactation. The objective of this study was to identify those QTL whose effects on milk production traits change during lactation. Test-day records were available for eight milk production traits on 1,829 first-parity Holstein cows. On average, each cow had 10.7 test-day records. The lactation was divided in 26 lactation stages of 15 days each. After SNP filtering, 30,348 SNP remained for the GWAS. A GWAS specifically for SNP by lactation stage interaction was performed. Significance for the interaction term was determined based on permutation of the genotypes. Eleven genomic regions showing significant SNP by lactation stage interaction were identified indicating that the effects of those regions change during lactation. The region on BTA 14 containing the diacylglycerol O-acyltransferase 1 (DGAT1) K232A polymorphism showed highly significant SNP by lactation stage interaction on milk yield, lactose yield, protein content and fat content. Another strong interaction was detected for a region on BTA 19 on lactose content. Less pronounced but significant interaction signals were detected for chromosomal regions on BTA 3, 9, 10b, and 27 on protein content, for regions on BTA 4 and 16 on fat yield and for regions on BTA 10a, 11, and 23 on fat content. The estimated effects for the lead SNP in the identified regions showed that the interaction was due to changes in SNP effects either in early or late lactation. We hypothesize that different SNP effects in early lactation are due to the negative energy balance which affects fat metabolism. Deviating SNP effects in late lactation might be due to pregnancy. Our study demonstrated that a GWAS for genotype by lactation stage interaction can identify new genomic regions involved in milk production. This will help to elucidate the genetic and biological background of milk production. The performed methods can also be used for other longitudinal traits like body weight or egg production.