Uniform sampling error

I tried to run beta diversity analysis through the standard qiime 2 workflow and I got the following error

tests-MacBook-Pro-4:data mortonjt$ qiime diversity beta --i-table palmspaper_targeted_filtered_sampleids_matched.qza --p-metric braycurtis --o-distance-matrix palm_ms_bray --verbose
Traceback (most recent call last):
  File "/Users/mortonjt/miniconda3/envs/bio/lib/python3.5/site-packages/q2cli-2017.2.0-py3.5.egg/q2cli/commands.py", line 217, in __call__
    results = action(**arguments)
  File "<decorator-gen-145>", line 2, in beta
  File "/Users/mortonjt/miniconda3/envs/bio/lib/python3.5/site-packages/qiime2-2017.2.0-py3.5.egg/qiime2/sdk/action.py", line 151, in callable_wrapper
  File "/Users/mortonjt/miniconda3/envs/bio/lib/python3.5/site-packages/qiime2-2017.2.0-py3.5.egg/qiime2/core/type/signature.py", line 282, in check_types
    " %r." % (name, spec.qiime_type))
TypeError: Argument to input 'table' is not a subtype of FeatureTable[Frequency] % Properties(['uniform-sampling']).

A couple of questions

  1. Am I getting this error because qiime diversity beta expects rarified a FeatureTable
  2. If the above is true, Is it possible to relax / extend this to other normalization schemes such as total sum scaling? There are cases where rarefaction is computationally intractable and impracticable. For instance, running rarefaction on metabolomics data will completely kill the RAM, since these are very high abundances.

Almost, it needs a normalized table of some form, the best word we could come up with is uniform-sampling as short for unifrom sampling effort. It is really just acting as a guard to ensure you’ve done some kind of normalization.

Ideally it is already relaxed, so another normalization scheme would just append the same semantic property onto its output type.