Full length PacBio/ONT 16S data

Does anyone have suggestions on best practices for using full-length 16S from nanopore or pacbio data with qiime2? It seems to me that the more error rich data would need alternative settings, but it’s not clear anyone has already done it.

I note that there are some full-length SILVA based classifiers, but if anyone could point me in a good direction, I’d appreciate it.

Hey @timp0,

We haven’t tried using that kind of data with QIIME 2 yet. It’s something we’re pretty interested in though, so if anyone else has tried this we’d love to hear how it went!

It seems to me like you may have the best luck with q2-deblur for this, since it makes choices based on reference sequences. Although I’m not sure what will happen with particularly noisy data.

I’ll try it. If you guys are interested in some data we’d be happy to share.


Hi ebolyen,Im wondering that does current version 2017.10 QIIME support Pacbio raw data for import? Or just illumina?is it possible the version 2018.1 soppurt pscbio raw data?
Thank you : )

Hey @Coke_Lin,

We definitely don’t have anything that handles PacBio data yet. We don’t really have much collective experience with it either. That being said, QIIME 2 is super extensible so if you knew some developers who interested in adding the formats and any QC needed, it should be pretty straight-forward (and we’d be happy to answer questions along the way).

Also depending on what the data is (e.g. amplicon vs shotgun vs long single molecule) there may not necessarily be the “downstream” analysis available yet either. We’ve focused mostly on amplicon data so far (and will continue to up to 2018.1), but someday we’ll have support for other kinds.


My nanopore data in this case at least is amplicon (full-length 16S). We are going to be (at least) trying to run it through QIIME in the near future.


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