Assessing kinship detection: Single nucleotide polymorphism array density and estimator comparison in white-tailed deer

Single nucleotide polymorphism (SNP) arrays have become increasingly
popular due to their affordability, commercial availability, statistical
power, and reproducibility. These arrays are being developed commercially
for a wide range of species in various density formats. In this study, we
evaluated the ability of commercially available medium-density (72,732
SNPs) and high-density SNP (702,183 SNPs) array for white-tailed deer
(Odocoileus virginianus) to accurately identify known genetically related
individuals within a wild population. We also assessed the impact of SNP
filtering thresholds on relatedness analyses and compared the performance
of four common relatedness softwares: KING, COLONY, Sequoia, and
COANCESTRY, on these known related pairs. Our analysis revealed that the
medium-density array exhibited greater tolerance to filtering and lower
sensitivity to bioinformatic pipelines, making it a favorable balance
between cost, computational time, and statistical power for analyses such
as population structure. Additionally, we found that reducing missing
data, specifically by using a subset of 600 loci with no missing data,
combined with the relatedness estimator Sequoia (which allows the
inclusion of life history data), yielded the most computationally
efficient and accurate results. These findings offer valuable insights
into the optimal SNP array size, appropriate filtering thresholds, and the
most effective genetic relatedness methods for wildlife population
studies.