Sequencing & Pangenomes

Optimal SNP selection for parentage discovery in large genotyped populations

---

Abstract

Genotyping is a common practice in beef cattle breeding programs, offering the added benefit of discovering and verifying parentage, improving pedigree accuracy. As genotyped populations grow and popular sires produce thousands of progeny, a robust set of markers is essential to avoid incorrect parentage assignments. This study aimed to identify an optimal set of SNP markers for parentage discovery in the American Angus Association population, which includes over 1.9 million genotyped animals. Assuming accurate pedigree records and no DNA sampling or identification errors, genotypes should reveal one sire and one dam per animal. Identifying more than two parents suggests false positives, while identifying only one or none indicates false negatives. Two benchmark datasets were used: a small set of 10,000 animals to test SNP sets of varying sizes, and a larger set of 74,000 animals to evaluate exclusion thresholds and stress-test the selected SNP set. Excluding self-matches, all other animals were considered candidate parents. Parentage status was determined using only homozygous SNPs for both animals, with the percentage of opposing homozygotes serving as the exclusion threshold. Eight SNP sets of different sizes were evaluated. The first two were the ISAG 200 and ISAG 554. The remaining six sets (200, 554, 1K, 2K, 4K, and 13K SNPs) were selected by filtering for call rate >0.99 and minor allele frequency (MAF) >0.35. Five exclusion thresholds (1-10%) were tested using the 2K SNP set and the large benchmark. Thresholds between 1-3% yielded similar results, with 3% offering the best balance between false positives and false negatives. This threshold was used for all subsequent analyses. SNP sets of 1K to 13K performed similarly, correctly identifying both sire and dam for over 97% of animals in the small benchmark. Smaller sets showed more variability: 554 SNP sets ranged from 88.6% to 96.6%, while 200 SNP sets ranged from 29.2% to 83.1%. In the large benchmark, the 2K SNP set achieved 98.5% correct parentage identification. The poorer performance of the ISAG sets was primarily due to the allele frequency distribution in the population, where many SNPs had low MAF reducing their power for parentage discovery. Nevertheless, the ISAG 554 set outperformed both 200-SNP sets, demonstrating the benefit of a larger set. These results demonstrate that a 2K SNP set provides robust parentage discovery in the American Angus population. An exclusion threshold between 1-3% effectively balances false positives and negatives. The ability to investigate individual cases of pedigree or sampling errors, and the relative importance of avoiding each type of error, should guide how strict or relaxed the threshold is set.

Keywords: 2026

How to Cite:

Garcia, A., Ramos, P., Medeiros, R., Novo, L. & Retallick, K., (2026) “Optimal SNP selection for parentage discovery in large genotyped populations”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2281873. doi: https://doi.org/10.31274/wcgalp.23460

Rights: 1