Broad Institute

Paired-end low-coverage whole genome shotgun reads were generated from 7 breeds (Arabian, Andalusian, Akhal-teke, Quarterhorse, Icelandic horse, Standardbred and Thoroughbred - 100,000 each). The WGS reads were placed uniquely on the Equus1.0 Thoroughbred assembly and SNP detection was carried out by SSAHA-SNP. The validation rate for these SNPs is estimated at approximately 95%. A second set of SNPs were discovered by comparing reads in the WGA assembly. SNPs were defined as mismatch positions that had a base quality of >= 30 on both reads in a region that aligned without gaps, and with at most one additional mismatch in the ten flanking bases.

Both sets include 948,609 SNPs in total distributed across chromosomes 1-31 and X. SNPs falling on contigs that are not assigned to chromosomes are excluded from this set. http://www.broadinstitute.org/ftp/distribution/horse_snp_release/v2