joined reads before deblur

Hi :slight_smile:

I found your current protocol paper about qiime2 2019.10 version(QIIME 2 Enables Comprehensive
End-to-End Analysis of DiverseMicrobiome Data and Comparative Studies with Publicly Available Data. )And I try to learn steps-by-steps with this static tutorial. And I have a very quick question related to the joined-methods.

In the paper, you used single-end data, so there are no details about how to join pair-end reads.

Q1 : In my case, I have raw pair-end reads( with primer, without barcodes, demultiplexed). I also have joined-reads, fasta format. I guess if my joined-reads is without quality score, I couldn’t use it as input for deblur? Right? So I should join my raw pair-end reads first.

Q2 : I searched on the post and found that I could use q2-vsearch join-paris function. Like this linkage ( ). But there is no primer removal procedure here. In dada, there is parameter called trim-left-length for me to trim primers. So in that case, should I perform cutadapt to trim primer first? And then import the output to perform quality-filter and deblur denoise-16S?

Thanks very much!


Hi @Lu_Zhang,

Thanks for the reminder that I should update that to reflect the newest version of QIIME 2. Btw, you if you are using the newest 2020.6 release, to follow that protocol, you will have to change one thing, as per the notice here when you get to the core-metrics step you'll need to change the n_jobs parameter parameter to n_jobs_or_threads which was recently changed.

As per your other questions:

You couldn't use the q2-deblur plugin, but I believe the stand-alone deblur does allow fasta files.

If your plan is to use Deblur, then yes! If you use DADA2, you should not join prior as DADA2 will do merging on its own.

Yes and no. Technically, neither DADA2 or Deblur have specialized primer/adapter removal procedures, the way something like q2-cutadapt does, though both offer options to do a fixed trim from either direction which can be used to trim primers. Deblur does offer options to trim from both the 5' and 3'. --p-trim-length will cut from the 5', and --p-left-trim-len will trim from the 3'. However, I personally would recommend using q2-cuatadapt to remove primers prior to merging/filtering, and making sure to enable the --p-discard-untrimmed tag, which will also discard any reads that don't have the primers in them. I find this does a great job of filtering junk reads.

Hope this helps!

Hi :slight_smile: @Mehrbod_Estaki

It helps me a lot! Thanks so much!


1 Like

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