[Feature Classifier Error] Not enough value to unpack

Hi all,

I am currently training the classifier and encountered a problem very similar to that of another user here. However, when I looked at my files and try to resolve the problem using their solution, I still cannot solve it. What I did was as follows:

Extract the reference reads from Greengenes reference database:

qiime feature-classifier extract-reads \
--i-sequences 2022.10.seqs.fna.qza \
--p-f-primer ACTCCTACGGGAGGCAGCAG \ 
--p-r-primer GGACTACHVGGGTWTCTAAT \ 
--p-trunc-len 446 \ 
--p-min-length 100 \
--p-max-length 450 \
--o-reads reference-seqs.qza

Training the classifier:

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

And the problem returned is:

Plugin error from feature-classifier: 
not enough values to unpack (expected 2, got 0)

I tried to open the input files (i.e., the reference-seqs.qza and the ref-tax.qza) and see if the problem lies in the # character as in the post above, but I see none. Could you please assist me to solve this problem?

My files are a little bit too big, so I uploaded them here.

Thanks a lot for your assistance!! :smiling_face:

Hi @hoangle,

I do not think these files are correctly paired with one another. That is, it appears that none of the IDs match between the reference-seqs.qza and ref-tax.qza files. This is required to build a classifier.

Hi @SoilRotifer,

Oh silly me! I realized that I might have downloaded the reference taxonomy and the full length gene database from 2 different sources, hence the difference in the IDs. I have fixed the problem and things worked out just fine!!
Thanks so much for your help!

2 Likes