See this thread -- basically, this would mean being more conservative about filtering stuff out of the table if it isn't already in the metadata (requiring the user to specify a --filter-missing-samples flag or something to do so). This would also apply for filtering out features in the table that aren't in the rankings (requiring the user to specify a --filter-missing-features flag or something).
We could either filter missing samples/features (as is done now) or assign it "placeholder" data. In the context of Qurro I think filtering makes more sense, hence naming these flags --filter-* instead of --ignore-*. But up to debate.
The current behavior in Qurro is just filtering the table and metadata as needed, while Emperor and Empress are somewhat more cautious.
This would technically be a backwards-incompatible change, so worth updating documentation / being careful to make the new UI intuitive for users.
Based on conversation with @ElDeveloper (thanks!) 
See this thread -- basically, this would mean being more conservative about filtering stuff out of the table if it isn't already in the metadata (requiring the user to specify a
--filter-missing-samplesflag or something to do so). This would also apply for filtering out features in the table that aren't in the rankings (requiring the user to specify a--filter-missing-featuresflag or something).We could either filter missing samples/features (as is done now) or assign it "placeholder" data. In the context of Qurro I think filtering makes more sense, hence naming these flags
--filter-*instead of--ignore-*. But up to debate.The current behavior in Qurro is just filtering the table and metadata as needed, while Emperor and Empress are somewhat more cautious.
This would technically be a backwards-incompatible change, so worth updating documentation / being careful to make the new UI intuitive for users.
Based on conversation with @ElDeveloper (thanks!)