Fail to train feature classifier from EzBiocloud database

[RayLin]

Hi everyone, I am trying to train my own classifier following the steps from the Tutorials by using the database downloaded from Ezbiocloud, include taxonomy txt file and OTU sequence fasta file. The format of taxonomy file is not QIIME’s default so at first I modified them manually.
(Before: 81152 Bacteria;Thermotogae;Thermotogae_c;Thermotogales;Fervidobacteriaceae;A61579_g;A61579_s
81153 Bacteria;Proteobacteria;Alphaproteobacteria;Rhodospirillales;Rhodospirillaceae;Magnetospirillum;Magnetospirillum magnetotacticum)
(After: 81152 k__Bacteria; p__Thermotogae; c__Thermotogae; o__Thermotogales; f__Fervidobacteriaceae; g__A61579; s__A61579 81153 k__Bacteria; p__Proteobacteria; c__Alphaproteobacteria; o__Rhodospirillales; f__Rhodospirillaceae; g__Magnetospirillum; s__Magnetospirillum magnetotacticum)

It seems good until the last step " Test the classifier". It showed the massage "Plugin error from feature-classifier:

this classifier does not support confidence values

Debug info has been saved to /tmp/qiime2-q2cli-err-r87x189t.log"

Would you please help me to debug them. Thank you!

Running VirtualBox Qiime2 in the Windows 7. All commands were listed below.

qiime tools import
--type 'FeatureData[Sequence]'
--input-path ezbiocloud_qiime_full.fasta
--output-path ezbio.qza

qiime tools import
--type 'FeatureData[Taxonomy]'
--input-format HeaderlessTSVTaxonomyFormat
--input-path ezbiocloud_id_taxonomy.txt
--output-path ref-taxonomy.qza

qiime feature-classifier extract-reads
--i-sequences ezbio.qza
--p-f-primer GTGCCAGCMGCCGCGGTAA
--p-r-primer GGACTACHVGGGTWTCTAAT
--p-trunc-len 120
--p-min-length 100
--p-max-length 400
--o-reads ref-seqs.qza

qiime feature-classifier fit-classifier-naive-bayes
--i-reference-reads ref-seqs.qza
--i-reference-taxonomy ref-taxonomy.qza
--o-classifier classifier.qza

qiime feature-classifier classify-sklearn
--i-classifier classifier.qza
--i-reads rep-seqs.qza
--o-classification taxonomy.qza

[ChrisKeefe]

Welcome to the forum, @RayLin!
I've never run into this issue myself, but it does come up occasionally on this forum. The error is generally caused by incorrect formatting of the taxonomy, usually from an uneven number of taxonomic levels or incorrect punctuation/whitespace.

This topic has some useful diagnostic discussion and tools. You'll find other examples of people troubleshooting their errors by searching this classifier does not support confidence values

Let us know what you find!
Chris

[RayLin]

Hi Chris, I have resolved this problem, maybe something wrong when I was editing the taxonomy file. In fact the first error I met when I was trying to train the classifier by the original taxonomy file download from EzBiocloud is "unable to allocate 3.92 gib for an array with shape (64275, 8192)".

Adjust the memory of VirtualBox to 6 GB can easily resolve this error. So we don't have to modify the format from EzBio, just use higher memories.

Thank you!

Best regards,

Ray

[ChrisKeefe]

Glad to hear everything worked out! Thanks for letting us know.

CK