18S data Taxonomic classifiers

Dear all,
Now, I want to analysis the 18SV9 data.,But Idont know, which Taxonomic classifiers should I use?
“Silva 132 99% OTUs full-length sequences” or “Silva_132_release.zip”

What are differences between these classifiers? should I convert it to .qza?

Hi @YuZhang,

These two sets of files are quite different. If you would like to curate your own SILVA reference database then the files from SILVA (your first screen-shot) can be downloaded and imported into QIIME 2 for further processing.

We’ve made this part much easier for you, via the RESCRIPt plugin. :building_construction:

Otherwise, the pre-made classifiers (your second screen-shot), and the files used to make them, can be found on the Data resources page. These have already been preprocessed through RESCRIPt for you. :package:


1 Like

I’m not sure if I got that right. So, the “Silva 132 99% OTUs full-length sequences” is the result of the “Silva_132_release.zip” after “RESCRIPt plugin” process ? So, if I want use Silva_132_release.zip, I could use the “Silva 132 99% OTUs full-length sequences” ( second screen-shot) directly. If Iwant use other reference database,I should run [RESCRIPt plugin] first, right? And can “Silva 132 99% OTUs full-length sequences” be an 18s classifiers? And I used “99_classifier.qza” to classify ITS, and 97_classifier.qza to classify 16s, is it right?

Sorry, let me clarify. But no, RESCRIPt was not implemented until QIIME 2 version 2020.6, and we’ve only provided pre-made files and classifiers for the latest SILVA 138 database since then. But it is perfectly fine to use the pre-made SILVA 132 files as long as you are using the corresponding version of QIIME 2.

We released pre-made SILVA 132 classifiers for 2020.2 and earlier using a different processing approach, with slightly different taxonomy string formatting. That is, we’ve updated the taxonomy prefixes to resemble those of GreenGenes (i.e. p__, c__, etc.) as of 2020.6. So, we only provide pre-made SILVA 138 classifiers and files.

However, you can use the RESCRIPt command qiime rescript get-silva-data (use the --help flag to read the help text) to grab version 132 and process / format to your liking, following the tutorial I linked. Note: RESCIPt currently requires QIIME 2 version 2020.8 or later.

Yes, the SILVA database can be used for both 18S and 16S data. Links to the ITS data files are linked in the Data Resources page. You’ll likely have to download and import those yourself.

Thanks alot.

Im not sure
“pre-made SILVA 132 classifiers for 2020.2 and earlier” is “Silva 132 99% OTUs full-length sequences” or not ?
My QIIME2 is 2018.11 version, so if I can use it to analysis my 18s data?

If I have the 2020.6 version Qiime2, I can use pre-made SILVA 138 classifiers ,or use RESCRIPt process other reference database,right?

All of the QIIME 2 documentation is archived for each version. So, if you are using 2020.6 then you can download the pre-made classifiers for 2020.6. That is if you go to the QIIME 2 Documentation, you’ll see a drop-menu for the version you have:


Or you can simply change the version number in the address bar of your browser, i.e. can change https://docs.qiime2.org/2020.11 to https://docs.qiime2.org/2020.6.

So, in your case go to https://docs.qiime2.org/2020.6/data-resources/ and you can use any files there. Again, you must have version 2020.8 to use RESCRIPt.


1 Like

Sir, I used the Version:2020.8.11, then I download

to classisfied my 18s data.
I ran the command below:

qiime feature-classifier classify-sklearn \
--i-classifier /share/silva-132-99-nb-classifier.qza \
--i-reads ../Analysis/core_data/rep-seq-dada2-1.qza \
--o-classification ../Analysis/core_data/taxonomy.qza \

qiime metadata tabulate \
--m-input-file ../Analysis/core_data/taxonomy.qza \
--o-visualization  ../Analysis/core_data/taxonomy.qzv \

qiime taxa barplot \
--i-table ../Analysis/core_data/table-dada2.qza \
--i-taxonomy ../Analysis/core_data/taxonomy.qza \
--m-metadata-file ../Analysis/core_data/metadata.tsv \
--o-visualization ../Analysis/core_data/taxa-bar-plots.qzv \

but there is a error:

I dont know how to do. Could you tell me where is run?

I am wondering if you are using the incorrect files. I noticed that you passed in a sequence file labeled rep-seq-dada2-1.qza

Yet your table is named table-dada2.qza. Assuming you are keeping your files similarly named I would expect you to have another file named as table-dada2-1.qza. Make sure you are using the correct set of matching files.

If this is not the case, can you share your taxonomy.qza and table-dada2.qza file?


Thanks alot. I do a wrong file. It should be table-dada2-1.Thanks a lot.!

1 Like