This study aimed to determine the effect of different rates of marker genotyping error on the accuracy of genomic prediction that was examined under distinct marker and quantitative trait loci (QTL) densities and different heritability estimates using a stochastic simulation approach. For each scenario of simulation, a reference population with phenotypic and genotypic records and a validation population with only genotypic records were considered. Marker effects were estimated in the reference population, and then their genotypic records were used to predict genomic breeding values in the validation population. The prediction accuracy was calculated as the correlation between estimated and true breeding values. The prediction bias was examined by computing the regression of true genomic breeding value on estimated genomic breeding value. The accuracy of the genomic evaluation was the highest in a scenario with no marker genotyping error and varied from 0.731 to 0.934. The accuracy of the genomic evaluation was the lowest in a scenario with marker genotyping error equal to 20% and changed from 0.517 to 0.762. The unbiased regression coefficients of true genomic breeding value on estimated genomic breeding value were obtained in the reference and validation populations when the rate of marker genotyping error was equal to zero. The results showed that marker genotyping error can reduce the accuracy of genomic evaluations. Moreover, marker genotyping error can provide biased estimates of genomic breeding values. Therefore, for obtaining accurate results it is recommended to minimize the marker genotyping errors to zero in genomic evaluation programs.

这项研究旨在确定不同的标记基因型分型错误率对基因组预测准确性的影响，在随机标记方法和定量性状基因座（QTL）密度和不同遗传力估计下，使用随机模拟方法对基因组预测准确性进行了检查。对于每种模拟情况，都考虑了具有表型和基因型记录的参考人群和仅具有基因型记录的验证人群。在参考种群中评估标记效应，然后将其基因型记录用于预测验证种群中的基因组育种值。计算预测准确性，作为估计育种值与真实育种值之间的相关性。通过计算真实基因组育种值对估计的基因组育种值的回归来检验预测偏差。在没有标记基因分型错误的情况下，基因组评估的准确性最高，介于0.731至0.934之间。在标记基因分型误差等于20％的情况下，基因组评估的准确性最低，从0.517变为0.762。当标记基因分型错误率等于零时，在参考人群和验证人群中获得了真实基因组育种值对估计基因组育种值的无偏回归系数。结果表明，标记基因分型错误会降低基因组评估的准确性。此外，标记的基因分型错误可以提供对基因组育种值的偏倚估计。因此，为了获得准确的结果，建议在基因组评估程序中将标记的基因分型错误最小化为零。