Effect of Accuracy of QTL Parameters Estimation on Genetic Response to Genotype Assisted Selection in Farm Animals

Abstract

Selection on estimated QTL effects could be detrimental if the estimated effect is inaccurate. In this study, effect of inaccurate estimation of QTL effect on Genotype Assisted selection (GAS) was investigated. We assumed various estimated QTL effects and various associated standard errors, and further, varied heritability and initial QTL allele frequency. True QTL effects were sampled from the posterior of the likelihood of QTL-hat and a prior gamma distribution of QTL effects. The ratio of GAS response and non-GAS (mass selection) response over repeated samples for a true QTL effect were averaged. The results showed that the GAS/non-GAS ratios was >1.267 for QTL effects of 0.5 of one phenotypic standard deviation (σp) when the SE was < 0.02 and heritability 0.2. For an effect of 0.4 and 0.3, the SE should be less than 0.141 and 0.117, respectively, before MAS becomes profitable. Assuming no error of QTL effect would give 10.61%, 5.34% and 2.31% more gain for effects of 0.4, 0.3 and 0.2σp respectively. However, accounting for a standard error of 0.05 would give estimated benefits of MAS over non-GAS of 8.89%, 4.35% and 1.56% respectively.