No unassigned taxonomy percentage


(Meriç Kınalı) #1

Hi all,

I have 16S illumina paired end sequences. Produced rep_seqs using dada2, import and then used ncbi 16sMicrobial.fasta and taxonomy files as a reference.
I was wondering that why I cant see any unassigned taxonomy percentages in the taxa_bar_plot of my samples, does this mean it can assign 100%, is this normal?
I followed steps from

qiime feature-classifier extract-reads
–i-sequences ncbi_fasta.qza
–o-reads classifier_ref-seqs.qza

qiime feature-classifier fit-classifier-naive-bayes
–i-reference-reads classifier_ref-seqs.qza
–i-reference-taxonomy ncbi_taxonomy.qza
–o-classifier classifier.qza

qiime feature-classifier classify-sklearn
–i-classifier classifier.qza
–i-reads rep-seqs-dada2.qza
–o-classification taxonomy.qza
–p-confidence -1 --p-read-orientation same

qiime taxa barplot
–i-table dada2_table.qza
–i-taxonomy taxonomy.qza
–m-metadata-file metafile.tsv
–o-visualization taxa-bar-plots.qzv.

taxa-bar-plots.qzv (346.0 KB)

Thank you so much for the help :space_invader:,

(Nicholas Bokulich) #2

This could be normal — it all depends on what sorts of samples you are sequencing and how atypical they are.

HOWEVER, it is incorrect in your case. This is because of this:

From the help docs:

  --p-confidence FLOAT            Confidence threshold for limiting taxonomic
                                  depth. Provide -1 to disable confidence
                                  calculation, or 0 to calculate confidence
                                  but not apply it to limit the taxonomic
                                  depth of the assignments.  [default: 0.7]

You are disabling confidence calculation, meaning you are telling the classifier to just pick the top hit! So you are overclassifying. See the citation for q2-feature-classifier for more details.

(system) closed #3

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