All features ids must be present in taxonomy, but the following feature ids are not:" and then many ID, like b47db366a25679dd507c550a23b82ace, 805a72c6085c867f05de73719aaf8870, 0367367b5700726abc795a11b382c763, b8ea2a4a5dae7de827152b8c21f643b1.
No, you don’t need to necessary train a new classifier - proceed to qiime feature-classifier classify-sklearn, then use the output from that (FeatureData[Taxonomy]) as the taxonomy input to qiime taxa filter-table.
Edit: I forgot that your original issue was related to labels in the taxonomy — yes, you will need to retrain, then reclassify, then you can filter.
I tried the potential solution (i.e., 1. qiime feature-classifier fit-classifier-naive-bayes; 2. qiime feature-classifier classify-sklearn; 3. qiime taxa filter-table) and the error is now at “qiime feature-classifier classify-sklearn.”
For qiime feature-classifier fit-classifier-naive-bayes [which worked], I did:
qiime feature-classifier fit-classifier-naive-bayes
For qiime feature-classifier classify-sklearn [which did not work], I did:
qiime feature-classifier classify-sklearn
–i-reads “Rep Seq file post pair, QC, and deblur”
–o-classification “Assign Taxa file post pair, QC, and deblur”
The error for the latter command was:
“Plugin error from feature-classifier:
pop from empty list”