Abstract
Determining the appropriate sample size (N) for bulk RNA sequencing experiments is critical for obtaining reliable results. We show in two N = 30 profiling studies, comparing wild-type mice and mice in which one copy of a gene has been deleted, the N required to minimize false positives and maximize true discoveries found in the N = 30 experiment. Results from experiments with N = 4 or less are shown to be highly misleading, given the high false positive rate and the lack of discovery of genes later found with higher N. For a cut-off of 2-fold expression differences, we find an N of 6-7 mice is required to consistently decrease the false positive rate to below 50%, and the detection sensitivity to above 50%. More is always better for both metrics - and an N of 8-12 is significantly better in recapitulating the full experiment.A common way to reduce the false discovery rate in underpowered experiments is to raise the fold cutoff. We show that this strategy is no substitute for increasing the N of the experiment: it results in consistently inflated effect sizes and causes a substantial drop in sensitivity of detection. These data should be helpful to scientists in choosing their Ns.