Permutations on SKAT-O

[djenikappel]

Dear @zhanxw,

I have been working with rvtests for the past few months and it works well on my large scale WES.

I have a question regarding permutations on SKAT-O. I have tried to run the SKAT-O kernel with permutations and it does not work. The message saying how the statistical significance is being calculated only includes the weights, but not permutation. --kernel skato[nPerm=1000:alpha=0.05:beta1=1:beta2=25] SKAT-O test significance will be evaluated using weight = Beta[beta1 = 1.00, beta2 = 25.00]

I have tested the same command on SKAT and then it seems to work. --kernel "skat[nPerm=1000:alpha=0.05:beta1=1:beta2=25] SKAT test significance will be evaluated using 1000 permutations at alpha = 0.05 weight = Beta[beta1 = 1.00, beta2 = 25.00]

Is there any reason why permutations are not allowed in the SKAT-O algorhitm?

Thanks

[zhanxw]

The permutation test for SKAT-O has not implemented. Shall we put that on our todo-list? @dajiangliu

[bibb]

Hi, I wanted to know if the permutation feature will be implemented in the near future? I'm looking forward to that.

[jiyue1214]

Dear @zhanxw, I am enjoying using the rvtest for my analysis. I found the permutation question has been posted in 2018. After I tried rvtests/2.1.0, I found permutation still cannot be applied to SKATO or any burden test model if I using --kernel skato[nPerm=1000000:alpha=0.001:beta1=1:beta2=20]. Could I know whether the feature has not been implemented or, I did not use it in an appropriate way? Thanks a lot for your help.