Why in LEfSe (KW = 1 , Wilcox = 0.05) output P values with > 0.05 are also recorded?

Downstream analysis and statistics LEfSe
1

Hi @sma , I have two sets of samples (control and test). I want to run only Wilcoxon test (not kruskal_wallis test). Thus, I set Kruskal-Wallis 1.00 and Wilcoxon 0.05. But, I am seeing P value > 0.05 is also recorded in the lefse output table. Can anyone please tell me why?

Thanks
dc7

2

Hi -
The output table records only the K-W p-values. Since alpha level for this was set to 1 in your run, you could still see those > 0.05 in the output.
Best,
Siyuan

1 Like
3

Is there a way to run only Wilcoxon and record its P-value (< 0.05)?

Thanks,
DC7

4

Hi -

Apologies, but I don’t think it’s possible at the moment. LEfSe’s design philosophy is to use K-W as the initial test, and between subgroup Wilcoxon as the optional filtering. So it wouldn’t make sense to run Wilcoxon without K-W.

Thanks,
Siyuan

1 Like