Table 1.
Name | Outcome type | Computational P-values | Omnibus test | R 2 and KRV | Reference |
---|---|---|---|---|---|
MiRKAT | Continuous, binary | Yes (Davies) | Yes (MinP) | Yes | Zhao et al. (2015) |
MiRKAT-S | Time-to-event | Yes (Davies) | Yes (MinP)a | Yes | Plantinga et al. (2017) |
MMiRKAT | Multivariate | Yes (Davies) | No; use KRV omnibus | Yes | Zhan et al. (2017b) |
KRV | Structured high-dimensional | Yes (Moment matching) | Yes (Omnibus kernel) | Yes | Zhan et al. (2017a) |
MiRKAT-R | Continuous; robust regression | Yes (Moment matching) | Yes (Omnibus kernel) | Yes | Unpublished |
CSKAT | Correlated continuous | Yes (Davies) | Yes (MinP) | No | Zhan et al. (2018) |
GLMM-MiRKAT | Correlated continuous, binary or Poisson | No | Yes (MinP) | No | Koh et al. (2019) |
Note: Tests with computational P-value calculation often also provide permutation P-values, which may be preferred for small samples.
Introduced in Koh et al. (2018).