A BOOTSTRAP APPROACH TO FROM CONFIDENCE-REGIONS FOR GENETIC-PARAMETERS METHOD-R ESTIMATES

Confidence regions (CR) for heritability (h(2)) and fraction of varian ce accounted for by permanent environmental effects (c(2)) from Method R estimates were obtained from simulated data using a univariate, rep eated measures, full animal model, with 50% subsampling. Bootstrapping techniques were explored to assess the optimum number of subsamples n eeded to compute Method R estimates of h2 and c2 with properties simil ar to those of exact estimators. One thousand estimates of each parame ter set were used to obtain 90, 95, and 99% CR in four data sets inclu ding 2,500 animals with four measurements each. Two approaches were ex plored to assess CR accuracy: a parametric approach assuming bivariate normality of h2 and c2 and a nonparametric approach based on the sum of squared rank deviations. Accuracy of CR was assessed by the average loss of confidence (LOSS) by number of estimates sampled (NUMEST). Fo r NUMEST = 5, bootstrap estimates of h(2) and c(2) were within 10(-3) of the asymptotic ones. The same degree of convergence in the estimate s of SE was achieved with NUMEST = 20. Correlation between estimates o f h(2) and c(2) ranged from -.83 to -.98. At NUMEST < 10, the nonparam etric CR were more accurate than parametric CR. However, with the para metric CR, LOSS approached zero at rate NUMEST-1. This rate was an ord er of magnitude larger for the nonparametric CR. These results suggest ed that when the computational burden of estimating genetic parameters limits the number of Method R estimates that can be obtained to, say, 10 or 20, reliable CR can still be obtained by processing Method R es timates through bootstrapping techniques.