The part I could not understand in ANCOM

I was trying to find out the features differ in abundance across different groups with ANCOM following “Moving picture tutorial”.

ANCOM assumes that few (less than about 25%) of the features are changing between groups. If you expect that more features are changing between your groups, you should not use ANCOM as it will be more error-prone (an increase in both Type I and Type II errors is possible).

I would like to know the details of this one sentence because I couldn’t
understand what it meant.

If someone would be so kind , could you please explain it clearly ?

Thank you.

1 Like

As I understood, if in your dataset abundances of more than 25% of features are changing between tested groups, it will affect theory testing. In one case, the correct null theory (no differences) may be rejected for some features, in other - test may fail to detect existing significant differences in features abundances. Or, false discovery rate increases and statistical power decreases.
But I have some doubts as well about this part:

Are they mean presence/absence differences or differences in abundances of only presented in all groups features :thinking: .