qiime feature-classifier extract-reads only forward primer

Hello QIIME2 Team!

I am working with forward reads (V3-V4). I just uploaded representatives sequences on fasta format into QIIME (they are all 265b of long) and now I want to classify the sequences against a particular database. Therefore, first I am trying to cut the database sequences using only the V3-V4 forward primer but it seems that it is not posible:

qiime feature-classifier extract-reads
--i-sequences data_base.qza
--p-min-length 260
--p-max-length 270
--o-reads database_cut.qza

There was a problem with the command:
(1/1) Missing option '--p-r-primer'.

I thought of using a conserve sequence between V3 and V4 as "primer" for the cutting, however would it give me a very short sequence (approx. 150b) losing around of 100b.

Do I have any option to include most of the forward read in the cutting.
Thank you very much for your help and your hard work always.


Just to clarify, are you are wanting to use your representative sequences to train the classifier? If so, you should be able to hand the .qza containing those directly to feature-classifier fit-classifier!

1 Like

Sorry for the misunderstanding. I mean, I want to use the database to train the classifier. But since my representative sequences are just forward I don't have a "reverse primer" so use to cut the database sequences to train the classifier. I hope this clarifies it, thanks!


Ah got it. Having the output from extract-reads exactly match your sequences is not important, so instead you can just hand it a generic reverse primer that is past the end of your reads, such as 806R(GGACTACVSGGGTATCTAAT), then use --p-trunc-len to cut to length.

Hope this helps!

1 Like

Oh thank you very much Keegan! I really appreciate your help!

1 Like

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