The present research aimed to characterize the Persian Kurdish horse population relative to the Persian Arabian and American Thoroughbred populations using genome-wide SNP data. Fifty-eight Kurdish, 38 Persian Arabian and 83 Thoroughbred horses were genotyped across 670,796 markers. After quality control and pruning to eliminate linkage disequilibrium between loci which resulted in 13,554 SNPs in 52 Kurdish, 24 Persian Arabian and 58 Thoroughbred horses, the Kurdish horses were generally distinguished from the Persian Arabian samples by Principal Component Analyses, cluster analyses and calculation of pairwise FST. Both Persian breeds were discriminated from the Thoroughbred. Pairwise FST between the two Persian samples (0.013) was significantly greater than zero and several fold less than those found between the Thoroughbred and Kurdish (0.052) or Thoroughbred and Persian Arabian (0.057). Cluster analysis assuming three genetic clusters assigned the Kurdish horse and Thoroughbred to distinct clusters (0.942 in cluster 2 and 0.953 in cluster 3 respectively); the Persian Arabian was not in a distinct cluster (0.519 in cluster 1), demonstrating shared ancestry or recent admixture with the Kurdish breed. Diversity as quantified by expected heterozygosity was the highest in the Kurdish horse (0.342), followed by the Persian Arabian (0.328) and the Thoroughbred (0.326). Analysis of Molecular Variance showed that 4.47% of the genetic variation was present among populations (P< 0.001). Population-specific inbreeding indices (FIS) were not significantly different from zero in any of the populations. Analysis of individual inbreeding based on runs of homozygosity using a larger SNP set suggested greater diversity in both the Kurdish and Persian Arabian than in the Thoroughbred. These results have implications for developing conservation strategies to achieve sound breeding goals while maintaining genetic diversity.

Document Type


Publication Date


Notes/Citation Information

Published in PLOS One, v. 16, no. 2, e0247123.

© 2021 Yousefi-Mashouf et al.

This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Digital Object Identifier (DOI)


Funding Information

NYM and HMY, The fund was provided by a private company, Ava Tejarat Saba for non-profit purposes.

Related Content

The full SNP dataset in PED/MAP format has been deposited on the Open Science Framework website under DOI 10.17605/OSF.IO/TR9KW which can be accessed by the URL: https://osf.io/tr9kw/.

journal.pone.0247123.s001.docx (13 kB)
S1 Table. Summary of the total number of runs of homozygosity by each size class. https://doi.org/10.1371/journal.pone.0247123.s001

journal.pone.0247123.s002.docx (259 kB)
S2 Table. Results on the comparisons of K = 1 to 5 tested by STRUCTURE Harvester. https://doi.org/10.1371/journal.pone.0247123.s002

journal.pone.0247123.s003.docx (12 kB)
S3 Table. Average proportion of membership of each pre-defined population in each of the 3 clusters at K = 3. https://doi.org/10.1371/journal.pone.0247123.s003

journal.pone.0247123.s004.docx (15 kB)
S4 Table. Average proportion of membership of each pre-defined population (excluding the Thoroughbred samples from the dataset) in each of the 2 clusters at K = 2. https://doi.org/10.1371/journal.pone.0247123.s004