نوع مقاله : مقاله پژوهشی
نویسندگان
1 دانشگاه کشاورزی و منابع طبیعی ساری، دانشکده علوم دامی و شیلات، گروه علوم دامی
2 دانشگاه علوم کشاورزی و منابع طبیعی ساری
3 متخصص ژنتیک آماری در شرکت آویاژن
4 دانشگاه کشاورزی و منابع طبیعی ساری، دانشکده علوم دامی و شیلات
5 گروه علوم دامی- مرکز تحقیق و توسعه کشاورزی و منابع طبیعی صفیآباد دزفول
چکیده
کلیدواژهها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
Background and objective: The most important factors affecting the accuracy of genomic prediction include the number of markers, statistical method, minor allele frequency (MAF), heritability, and genetic architecture, that each of their effects differs under the specific conditions. The aim of this study was to investigate the effect of adding noncausal SNPs to the statistical model on the accuracy of breeding values by considering different population factors and different genetic architectures.
Materials and methods: In the present study, 10 chromosomes with a length of 100 cM were simulated and on each chromosome 1000 SNP markers of two alleles were considered which were randomly distributed along the genome. In this study, phenotypic variance was assumed 1.0 and heritability was considered as 0.1, 0.2 and 0.5 according to traits with low, medium and high heritability respectively. The population numbers for this simulation were 1,000, 2,000 and 4,000 people and 100 QTLs was distributed on 10 chromosomes. Predictions were done by RKHS and BayesB models and mean comparisons were also made using the Tukey method.
Results: In all scenarios and with both methods, there was a significant difference between considering QTLs alone and with different percentages of noncausal SNPs in the model. However, in both methods, significant differences were observed between low percentages of noncausal markers and high percentages.
In most scenarios, for example, when there were 10% noncausal markers in the model, the predictions were more accurate than when 20%, 40%, 60% and 100%.
The accuracy of genomic predictions increased with the increasing of heritability and the number of individuals in the reference population. In the case of low heritability scenarios, the percentage of increment of predictive accuracy was more than higher heritability with increasing of reference population. The mean of accuracy of evaluations using BayesB method compared RKHS, was significantly different in fitting QTLs in the model and different ratios of noncausal markers in most scenarios except for a few cases. In general, in both methods, the best predictions were made when using the highest heritability and population size and only entering QTLs in the models.
Conclusion: According to these findings, identifying causal sites and incorporating them into genomic assessments increases the rate of analysis and reduces the cost of genotyping. It can also enhance the genetic gain of important economic traits in genomic evaluations by enhancing predictive accuracy.
Conclusion: According to these findings, identifying causal sites and incorporating them into genomic assessments increases the rate of analysis and reduces the cost of genotyping. It can also enhance the genetic gain of important economic traits in genomic evaluations by enhancing predictive accuracy.
کلیدواژهها [English]
)