نوع مقاله : مقاله پژوهشی
نویسندگان
دانشگاه علوم کشاورزی و منابع طبیعی ساری
چکیده
کلیدواژهها
عنوان مقاله [English]
نویسندگان [English]
Back ground and objectives: Identification of single nucleotide markers and different methods of genomic evaluation in the form of marker assisted selection at the genome level has led to considerable genetic progress in the economic traits of domestic animals. The success of genomic prediction is measured by its accuracy. Deterministic formulas determine the relationship between prediction accuracy and factors affecting prediction accuracy and therefore before running into genomic selection, it is possible to design an optimal program such as the appropriate size of the reference population to achieve the optimum level of selection accuracy. The aim of the present study was to evaluate the prediction of the accuracy of deterministic formulas and compare it with the accuracy of prediction of genomic breeding values in the simulated study.
Materials and methods: Four deterministic formulas including Daetwyler et al formula, Goddard formula, Goddard et al formula and Rabier et al formula were used to predict the accuracy of genomic evaluation in different genetic architectures including different levels for heritability, reference population size and number of independent chromosome segments. The ShinyGPAS program was used to compare and plot the accuracy of prediction. In order to compare the performance of deterministic formulas with the accuracy of predictions in the simulated population, population simulations were performed using QMSIM software. For this purpose, in genome simulation, three levels of heritability of 0.1, 0.3 and 0.5 and two levels of reference population size of 1000 and 2000 individuals were considered and estimation of genomic breeding values was performed using Bayesian method A and Bayesian B using BGLR package in R medium
Results: In low heritability, the highest prediction accuracy was observed in Goddard formula, which had the closest prediction accuracy (0.56) to the accuracy of genomic evaluation of simulated data estimated by Bayes A method (0.56). With moderate heritability (0.3), Goddard (0.74) and Rabier et al. (0.73) had the closest and most similarity to the accuracy of the simulated data. With population size increased from 1000 to 2000 individuals along with increasing heritability, the performance of deterministic formulas was closer to the accuracy estimated from simulation data by Bayesian methods and the most agreement was obtained in Goddard and Rabier methods. In the lower independent chromosome segments, the highest accuracy obtained by Rabier et al (0.860). With increasing chromosomal independent segments, the highest value of accuracy obtained by Goddard predictive formula.
Conclusion: The results showed that deterministic formulas have a good ability to predict the accuracy of genomic evaluation and their performance is linked to the genetic architecture. The results suggest that the predictions of accuracy, in general, using Goddard and Rabier formulas are more consistent with genomic estimation accuracy in the simulated data.
کلیدواژهها [English]
)