AIMS:

We aim to show the analytical gains from increased sample size, but also that relatively low numbers of samples can be used to locate strong genetic signals in the presence of perfect intermediate measurements (here response to taste test challenge). We also which to use a dietary measure from Food Frequency Questionnaire data (likely the consumption of cruciferous vegetables) to illustrate the systematically poor power (at even relatively large sample sizes) when using a poor or distal phenotypic measure.