Using triallelic SNPs for determining parentage in North American yak ( ) and estimating cattle ( ) introgression.

 Genetic testing for pedigree accuracy is critical for managing genetic diversity in North American (NA) yak ( ), a population expanded mostly from imported zoological park specimens.  DNA testing also enhances species conservation by identifying recent F1 hybrid ancestors (within three generations).  Biallelic single nucleotide polymorphisms (SNPs) can accomplish either task, but increases the marker count and costs necessary to achieve both.  Our aim was to identify novel, multifunctional, triallelic yak SNPs (tySNPs), with each having two alleles for yak parentage testing, and a third allele for identifying recent cattle introgression.    Genome sequences were aligned to the cattle UMD3.1 assembly and SNPs were screened for 1) heterozygosity in a NA and a Chinese yak, 2) a third allele at high frequency in cattle, and 3) flanking sequences conserved in both species.  Subsequently, tySNPs were filtered for unique alignment to the haplotype-resolved F1 yak assembly.  Allele frequencies were estimated in a subset of 87 tySNPs by genotyping 170 NA yak.   We identified 610 autosomal tySNPs, distributed in 441 clusters with 5 Mb average genome spacing.  The average NA yak minor allele frequency was high (0.296), while average introgressed cattle alleles were low (0.004).  In simulations with tySNPs, 28 were sufficient for globally-unique animal identification (P =5.81x10 ), 87 were able to exclude 19 random bulls from parentage at the 99% level without using the dam's genotype (P =5.3x10 ), and 87 were able to detect F1 hybridization events after three generations of yak backcrosses (1/16th germplasm). :  Identifying animals, determining parentage and detecting recent hybridization events was efficient with as few as 87 tySNPs.  A similar triallelic approach could be used with other bottlenecked species that hybridize with cattle, such as NA plains bison ( ).