correction via permutation is not robust to confounding and it is essential to assess the quality of the distribution of test statistics, for those genes that have i-stats less than the threshold, to ensure appropriate calibration of the distribution. Nevertheless, with increasing sample sizes, the dimensionality of the tests will also increase, and studies will be assessing close to 20,000 tests. Therefore, for large studies we consider the Bonferroni threshold to be preferable.
