Uses beyond RNA?

[seanken]

Hey,

Hope all is well! A two part question for you:

1. Is there any reason not to use TRADE on non-RNA-seq data? I would think it should be reasonable on anything that estimates, for each feature, an effect size and a standard error for that effect size with glm based regression (or even glmm regression), unless I am missing something?
2. Assuming the answer to number 1 is that it is ok to use TRADE for other data types, I wonder if you have any conception of how many features (genes, proteins, etc) one would need in a dataset to do so? If I recall, TRADE uses the gene to gene variability to estimate the underlying distribution, but that would get harder the less genes there are (or at least that is my intuition). Any idea on what the minimal number would be?

Sorry if the above are addressed in the paper and I just forgot/missed them when I skimmed it again today.

Best wishes and many thanks,

Sean