Power of an exome-sequencing study is limited by the amount of variation in a gene. Therefore, power is higher for genes with more variants, for example longer genes or genes in regions of elevated mutation rate. Additionally, genes in which most variants are causal are easier to identify than those in which few variants are causal. In individual candidate gene sequencing studies estimates of this proportion ranged from 30% to 70%21, 22, 26. Consequently, the effect size is not only a property of an individual variant, but rather a reflection of the distribution of effects coupled with how those effects are interpreted via the test. Some statistical tests explicitly account for differences in power when evaluating evidence of association81.
