From: Power and sample size calculations in the presence of phenotype errors for case/control genetic association studies
