Why were only a few taxonomies discovered?

[Wang_cs001632]

Our code:
qiime feature-classifier classify-sklearn \

--i-classifier silva-132-99-nb-classifier.qza
--i-reads rep-skin-dada2.qza
--o-classification taxonomy_skin.qza
Thanf for your attention.

[SoilRotifer (Mike Robeson)]

Hi @Wang_cs001632,

There could be several reasons for this, some are outlined here:

Possible Analysis Pipeline for Ion Torrent 16S Metagenomics Kit Data in QIIME2?

Unclassified taxa can occur for a few reasons with the vsearch-based classifier: there really is no good match. Seems unlikely if this is region-specific, but that could occur if the reference sequences have a jumble of different regions covered (e.g., different primers were used for sequencing those reference sequences). It looks like you are using Greengenes? So that is probably not the case, either. Can occur if --p-perc-identity is set too high… but the default is really low, so that’s p…

and here:

Train you classifier - taxonomy

Hi @Jasmine84, See this post. Depending on the length of your sequences, the primers you are using, etc, deeper than order level might be unreliable. On 250 bp 16S rRNA gene sequences, we can usually get genus level and some species with a good degree of reliability. The naive bayes classifier in QIIME2 is designed to be cautious and reliable, to avoid false-positive errors, but you can lower the confidence parameter to make it less cautious. If you have: only kingdom-level classification, …

Since you are making use of the SILVA 132 full-length sequence classifier, my first guess would be that your reads may be in the wrong direction. You can make use of RESCRIPt to orient your reads to the SILVA reference database with the qiime rescript orient-seqs command.

One last suggestion, the SILVA 138 classifiers, and the files used to make them are here.

-Best wishes!

[Wang_cs001632]

Thank you for your timely reply!
The dataset to be processed was like that (ENA Browser)

SRR11242604.fastq.gz
SRR11242604_1.fastq.gz
SRR11242604_2.fastq.gz
Should we just simply cat them into one fastq.gz, such as
cat SRR11241832_1.fastq.gz SRR11241832_2.fastq.gz > 72S_L001_R1_001.fastq.gz
then input qiime2 as "CasavaOneEightSingleLanePerSampleDirFmt" ?

[SoilRotifer (Mike Robeson)]

I think it’d be easiest to do this after you’ve merged the reads and made ASVs. That is, run this command on your DADA2 or deblur output prior to assigning taxonomy:

qiime rescript orient-seqs ...

-Mike