A genome-wide association study of calf birth weight in Holstein cattle using single nucleotide polymorphisms and phenotypes predicted from auxiliary traits.

Previous research has found that a quantitative trait locus exists affecting calving and conformation traits on Bos taurus autosome 18 that may be related to increased calf birth weights, which are not routinely recorded in the United States. Birth weight data from large, intensively managed dairies in eastern Germany with management systems similar to those commonly found in the United States were used to develop a selection index predictor for predicted transmitting ability (PTA) of birth weight. The predictor included body depth, rump width, sire calving ease, sire gestation length, sire stillbirth, stature, and strength. Genetic and phenotypic correlations and heritabilities from the United States were substituted for the German values, and birth weight PTA predicted for 31,984 bulls with US genetic evaluations. A genome-wide association study was conducted on the predicted birth weight PTA with the 2-step genomic BLUP procedure used for routine evaluations in the United States. Allele substitution effects were predicted for 43,188 single nucleotide polymorphisms (SNP). Genotypes were available for 53,644 predictor animals. Gene set enrichment analysis was performed on the 100 SNP that had the largest effects expressed in additive genetic standard deviations. Several SNP related to growth and development were found among the 25 SNP with the largest effects, including markers located within or near (≤ 100 kbp) ABCA12, FLRT2, LHX4, MAP3K5, NRAC, NTNG1, PIGN, and ZNF75A. The gene set enrichment analysis identified the Kyoto Encyclopedia of Genes and Genomes "Regulation of actin cytoskeleton" pathway (bta04810) as being enriched. That pathway includes the ROCK gene, which is involved in placental function in the human, as well as other developmental genes (e.g., FAK and PAK). Prediction equations derived from one population are useful for identifying genes and gene networks associated with phenotypes that are not directly measured in a second population. This approach will identify only genes associated with the traits used to construct the birth weight predictor, and not loci that affect only birth weight.