Taxonomic Classifier Warning

[uth]

Hi,

I’m trying to run classification for sequences with silva database. I used the latest silva files downloaded from Archive

first I used qiime version 2019.10 and received the following error
“The TaxonomicClassifier artifact that results from this method was trained using scikit-learn version 0.21.2. It cannot be used with other versions of scikit-learn. (While the classifier may complete successfully, the results will be unreliable.)”

Since it says the results can be unreliable, I downloaded the latest version of qiime2 (2021.2) and ran all the 4 codes needed to make the classifier.

1. qiime tools import --type ‘FeatureData[Sequence]’ --input-path SILVA_132_QIIME_release/rep_set/rep_set_16S_only/99/silva_132_99_16S.fna --output-path silva_132_16S.qza

2. qiime tools import --type ‘FeatureData[Taxonomy]’ --input-format HeaderlessTSVTaxonomyFormat
   –input-path SILVA_132_QIIME_release/taxonomy/16S_only/99/taxonomy_7_levels.txt --output-path 16s-taxonomy-silva.qza

3. qiime feature-classifier extract-reads --i-sequences silva_132_16S.qza --p-f-primer AGGATTAGATACCCTGGTA --p-r-primer CRRCACGAGCTGACGAC --o-reads Silva_132_v5v6_ref.qza --verbose

4. qiime feature-classifier fit-classifier-naive-bayes --i-reference-reads Silva_132_v5v6_ref.qza --i-reference-taxonomy 16s_taxonomy_silva.qza --o-classifier silva_v5v6_classifier.qza --verbose

Unfortunately, I still get a warning message as below:

UserWarning: The TaxonomicClassifier artifact that results from this method was trained using scikit-learn version 0.23.1. It cannot be used with other versions of scikit-learn. (While the classifier may complete successfully, the results will be unreliable.)

Could anyone please help me understand what’s happening.

Many thanks in advance,
Uth

[thermokarst]

Hi @uth!

This is all indication that it is working as expected! The "warning" here is just letting you know, "hey, you made this classifier using this version of scikit-learn - don't mix and match with other versions of scikit-learn, or else you might produce invalid results."

Looks like you're all set - now its time to take that classifier out for a test drive!

Keep is posted on how it goes!

[uth]

Hi @thermokarst,

Thanks for the prompt response. Such a relief to hear that all is good to go. Will do the classification an d will post here if something goes wrong.

Many thanks again…

Uth