Genome-wide across and within breed association studies in horses

The objectives of this work were genome-wide association studies for equine osteochondrosis (OC) using the Illumina Equine 50K Beadchip in Hanoverian warmblood. In addition to these within-breed analyses, we performed across-breed analyses for breed susceptibility to OC including horse breeds known for being at very low risk to OC. From a large sample of Hanoverian warmblood horses, 244 most distantly related animals were chosen to avoid stratification of data by families. One hundred and fifty six animals were affected by OC in fetlock and/or hock joints and 88 were free from any signs of OC in each limb joint. Phenotypic traits were OC in fetlock and/or hock joints, osteochondrosis dissecans in fetlock and/or hock joints. These OC conditions were also analysed separately for fetlock and hock joints. Across-breed analyses included Hanoverian, Arabian, Black Forest, Icelandic horses and other pony breeds. For genome-wide mapping, we performed association analyses (MAF>0.05) controlling for cryptic structure of the genotypic data, non-genetic fixed effects and genomic relationships among animals. In Hanoverians, associated loci explained more than 70-85% of the phenotypic variance and for all these loci candidate genes were identified. The most significant hits for OC were on ECA1, 2, 5, 8, 9, 10, 12, 14, 16, 18, 20, 23 and 26. A number of these loci could be verified using across-breed mapping. We used these mapping results to develop an efficient strategy for selecting a minimum number of animals for next generation sequencing to capture the genetic variation causing OC as complete as possible.