Genome-wide Autozygosity Reveals Positive Selection Signatures in Racehorses

Artificial selection for elite racing performance is expected to leave distinct selection signatures on the equine genome. These signatures may differ between breeds due to individual racing performance characteristics (e.g., trotting vs. galloping), or there may be evidence of convergent selection for common functional variants associated with maximal exertional efforts at speed. Here, we investigated the distribution of autozygosity levels based on runs of homozygosity (ROH) in 510 Thoroughbreds, 328 Standardbred trotters, 147 Standardbred pacers, and 24 racing Quarter Horses genotyped for over 138,000 autosomal SNPs, and adjusted for population structure. Within each group, genomic regions of interest (ROI) were defined as those in which the number of animals in a ROH exceeded the upper outlier threshold with a magnifier of 3. Across breeds, 54 ROI were observed over 14 autosomes, ranging from 35kb to 2.2Mb. BioMart queries of ROI returned 465 transcripts, of which 314 were annotated. Transcripts unique to each group numbered 8, 15, 16, and 15 for Thoroughbred, Standardbred trotter, Standardbred pacer, and racing Quarter Horse, respectively. Thirty transcripts were common between Thoroughbred and Quarter Horse, 29 transcripts were shared between Standardbred trotters and pacers, and two transcripts were shared among Quarter Horse and both Standardbred groups. None were shared across all four groups, although ROI peaks in one group were often echoed by sub-threshold peaks in others. Subsequent steps include comparison with non-racing breeds and pathway analysis to identify protein-coding genes or variants associated with racing performance.