I’m not sure if I understand when to use median_comparison_abundance=FALSE. The manual says:
median_comparison_abundance should be FALSE when (1) testing for significant relative associations […]
As microbiota sequencing data is usually essentially relative (I mean, it does not really measure absolute abundances, even though the data has absolute counts), I would tend to think this is usually the case. Yet, the default is TRUE, which makes me think I misunderstand this somehow.
Copying this over from the other thread since I think it’s the same question:
“Relative” refers to the same thing you’re describing - you have that right. In the last few years, there’s been quite a bit of statistical work about trying to infer absolute abundance changes from relative abundance data given a variety of assumptions. Here, if you assume that fewer than half of the features are changing in absolute abundance and you have a large enough sample size and number of features (>100 samples and >200 features should probably be enough), the test against the median is equivalent to a test on the absolute scale. We hope to get a pre-print out soon that should explain this in more detail.
If you don’t want to worry about any of this and just test relative abundances, you can just set median_comparison_abundance to FALSE.