--p-trunc-len set in qiime feature-classifier extract-reads

Hi, community,

I am using 341F-CCTACGGGNGGCWGCAG, 785R-GACTACHVGGGTATCTAATCC, so I need to train my own classifiers.
Based on the tutorial https://docs.qiime2.org/2017.7/tutorials/feature-classifier/
I do not know how to set the --p-trunc-len is appropriate, is there anyone who can help me? What Is there a benchmark for setting the length?

And when I consider this question, I former used 515F-806R parimer and got paired ended sequences, is it appropriate to use the classifier in this website for 515F-806R paired ended sequences? https://docs.qiime2.org/2017.11/tutorials/moving-pictures/

Thanks so much.

I found the answer to the first question here https://github.com/qiime2/docs/issues/216
But I am still not sure about my second question. I always used the classifier used in the Moving Pictures tutorial. Just regarded it as single ended and paired ended all suitable primer (515-806).

1 Like

Hi @Lu_Yang,
Yes, looking at the provenance of that classifier file (you can use view.qiime2.org to check provenance of artifacts, in addition to viewing visualizations!), it seems that the classifier was trained on amplicons extracted with 515F/806R primer, without any further trimming. So this should be appropriate for single- or paired-end reads targeting this amplicon.

I hope that helps!

Hi, @Nicholas_Bokulich, really helpful! Thanks so much.


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