Table 1.
Predicted accuracy and percentage prediction error assessed by simulation with disease prevalence = 0.1 (SE range 0.0004–0.0065).
Table 2.
The effects of different distributions of allele frequency and effects on accuracy in a continuous phenotype with observed heritability = 0.5 (SE range 0.0004–0.0057).
Table 3.
Accuracy for continuous phenotype when setting 0.95 of nGa loci to zero (λ = 0.02 = 400nPb/20,000nG, SE range 0.0042–0.0057).
Figure 1.
Predicted accuracy of estimated genetic values of a continuous phenotype.
Predicted accuracy of estimated additive genetic values of a continuous phenotype as a function of observed heritability and number of phenotypes per genotype tested, λ = 0.02, 0.1, 0.5, 1, 2, 5, 10 and 20 from minimum to maximum accuracy respectively.
Table 4.
Accuracy for a dichotomous disease trait as prevalence varies (a, bλ = 1, SE range 0.0026–0.0048).
Table 5.
Simulated accuracy of a population study for a dichotomous phenotype as prevalence and a varies and
b stays constant (λc = 10,
, predicted accuracy = 0.816, Equation (4), SE range 0.0025–0.0038).
Figure 2.
Predicted accuracy of estimated genetic risk from population and case control designs of a dichotomous phenotype.
Contour plot of predicted accuracy for varied prevalence and additive heritability on the observed scale, in population studies (dashed vertical line) and case control studies (solid line) of dichotomous phenotypes. Each contour represents a line of constant accuracy, starting from the right 0.9, 0.8, 0.7, and 0.6. The narrowly dashed line is derived from Equation (5) with , so values below this line are not possible under the liability model.