fit classifier naives bayes Silva 138 database

Technical Support
,
1

Hi, I am having problems running the below script:

qiime feature-classifier fit-classifier-naive-bayes \

--i-reference-reads silva-138-99-seqs.qza
--i-reference-taxonomy silva-138-99-tax.qza
--o-classifier Silva_classifier.qza
--verbose

I get this warning message, not technically an error(?) but also no output:

/opt/conda/envs/qiime2-2021.4/lib/python3.8/site-packages/q2_feature_classifier/classifier.py:102: UserWarning: The TaxonomicClassifier artifact that results from this method was trained using scikit-learn version 0.24.1. It cannot be used with other versions of scikit-learn. (While the classifier may complete successfully, the results will be unreliable.)
warnings.warn(warning, UserWarning)

The silva database was downloaded from Qiime2 data resources as .qza files. As they are .qza do you still need to do the RESCRIPt process?

I have also seen on other posts about something similar being a memory issue or the incorrect version of scikit-learn. How do I find out the scikit-learn version I am using and if I do need update, how do I do that? I am using a HPC cluster with linux.

Thanks

Jess

2

Hi @JessG,

If you search the forum you'll find that the most common reason for encountering this error message is that you you are using a classifier that was constructed for a different version of QIIME 2.

So if you are still running qiime2-2021.4, then you need to use the classifier that was constructed for that version of QIIME 2, which is located at:

https://docs.qiime2.org/2021.4/data-resources/
Be sure to click the "Show me content on this page" button.

-Mike

2 Likes
3

Hello,

I am encountering a similar issue. I have downloaded pre-formatted SILVA reference sequence and taxonomy files (available at Data resources — QIIME 2 2023.7.0 documentation) for the version of QIIME2 that I am currently using. I extracted the reference sequences for the V4 region (515F/806R), and when training the classifier, I received the following message:

.../anaconda3/envs/qiime2-2023.7/lib/python3.8/site-packages/q2_feature_classifier/classifier.py:102: "UserWarning: The TaxonomicClassifier artifact that results from this method was trained using scikit-learn version 0.24.1. It cannot be used with other versions of scikit-learn. (While the classifier may complete successfully, the results will be unreliable.) warnings.warn(warning, UserWarning)"

I proceeded to test the classifier and visualize the taxonomic assignments, but I'm uncertain about the reliability of the results I obtained.

Thank you for all of your assistance.

4
5

Hi @Pam_mp,
If you search the forum you will see several instances where the error below is explained and resolved:

This post links to several other posts that should be helpful to you. Let me know if you have any questions from there.
-Hannah

3 Likes
6
7

This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.