Department of Animal Science, University of Nebraska, Lincoln, NE 68583, USA.
J Anim Sci. 2023 Jan 3;101. doi: 10.1093/jas/skad077.
Gene editing has the potential to expedite the rate of genetic gain for complex traits. However, changing nucleotides (i.e., QTN) in the genome can affect the additive genetic relationship among individuals and, consequently, impact genetic evaluations. Therefore, the objectives of this study were to estimate the impact of including gene-edited individuals in the genetic evaluation and investigate modeling strategies to mitigate potential errors. For that, a beef cattle population was simulated for nine generations (N = 13,100). Gene-edited sires (1, 25, or 50) were introduced in generation 8. The number of edited QTN was 1, 3, or 13. Genetic evaluations were performed using pedigree, genomic data, or a combination of both. Relationships were weighted based on the effect of the edited QTN. Comparisons were made using the accuracy, average absolute bias, and dispersion of the estimated breeding values (EBV). In general, the EBV of the first generation of progeny of gene-edited sires were associated with greater average absolute bias and overdispersion than the EBV of the progeny of non-gene-edited sires (P ≤ 0.001). Weighting the relationship matrices increased (P ≤ 0.001) the accuracy of EBV when the gene-edited sires were introduced by 3% and decreased (P ≤ 0.001) the average absolute bias and dispersion for the progeny of gene-edited sires. For the second generation of descendants of gene-edited sires, the absolute bias increased as the number of edited alleles increased; however, the rate of increase in absolute bias was 0.007 for each allele edited when the relationship matrices were weighted compared with 0.10 when the relationship matrices were not weighted. Overall, when gene-edited sires are included in genetic evaluations, error is introduced in the EBV, such that the EBV of progeny of gene-edited sires are underestimated. Hence, the progeny of gene-edited sires would be less likely to be selected to be parents of the next generation than what was expected based on their true genetic merit. Therefore, modeling strategies such as weighting the relationship matrices are essential to avoid incorrect selection decisions if animals that have been edited for QTN underlying complex traits are introduced into genetic evaluations.
基因编辑有可能加速复杂性状的遗传增益速度。然而，改变基因组中的核苷酸（即 QTN）会影响个体间的加性遗传关系，并因此影响遗传评估。因此，本研究的目的是估计在遗传评估中包含基因编辑个体的影响，并研究减轻潜在错误的建模策略。为此，模拟了一个牛群进行了九代（N = 13100）。在第八代引入了基因编辑的种公牛（1、25 或 50 头）。编辑的 QTN 数量为 1、3 或 13 个。遗传评估使用系谱、基因组数据或两者的组合进行。关系基于编辑 QTN 的效果进行加权。使用准确性、平均绝对偏差和估计育种值（EBV）的分散度进行比较。一般来说，与非基因编辑种公牛的后代相比，基因编辑种公牛的第一代后代的 EBV 与更大的平均绝对偏差和过度分散有关（P ≤ 0.001）。当引入 3%的基因编辑种公牛时，加权关系矩阵会增加（P ≤ 0.001）EBV 的准确性，并降低基因编辑种公牛后代的平均绝对偏差和分散度（P ≤ 0.001）。对于基因编辑种公牛的第二代后代，随着编辑等位基因数量的增加，绝对偏差增加；然而，当加权关系矩阵时，每个编辑等位基因的绝对偏差增加率为 0.007，而当不加权关系矩阵时为 0.10。总体而言，当在遗传评估中包含基因编辑的种公牛时，EBV 会引入误差，从而导致基因编辑种公牛的后代的 EBV 被低估。因此，与基于其真实遗传优势预期的相比，基因编辑种公牛的后代不太可能被选为下一代的亲本。因此，如果引入了针对复杂性状的 QTN 进行编辑的动物，那么诸如加权关系矩阵等建模策略对于避免错误的选择决策至关重要。
