Zhou, ZhengyangKu, Hung-ChihXing, Chao2020-12-122020-12-122020https://hdl.handle.net/20.500.12503/30246Purpose: In genetic association studies with case-control design, standard practice is to perform the Cochran-Armitage (CA) trend test under the assumption of additive genetic model. The CA trend test is a parametric statistical test, and under the null hypothesis of no association between the genetic variant and disease, the test statistic asymptotically follows a chi-square distribution with 1 degree-of-freedom. However, when the sample size and/or variant minor allele frequency are small, asymptotic properties may not hold, which can lead to reduced statistical power in detecting genetic associations. Methods: To improve statistical power in this case, we consider the two-sided Jonckheere's test, which is a rank-based nonparametric test. By not imposing assumptions on the distributions of the data, it is expected to have better statistical power than parametric tests for small sample sizes and/or rare variants. We conducted extensive simulations to compare the statistical power between the CA trend test and the two-sided Jonckheere's test under various scenarios. Results: We found for small sample size (e.g., n=200) and low minor allele frequency (e.g., p=0.05), the two-sided Jonckheere's test outpowered the CA trend test for all genetic models ranging from recessive to dominant. Conclusion: This finding provides an alternative to the CA trend test in genetic association studies under these circumstances. With higher statistical power from the two-sided Jonckheere's test, genetic epidemiologists will be able to detect more genetic associations for complex diseases, which may lead to better prevention and treatment strategies.enposter