Merging paired reads

Hi Greg, et. al.,

I was rarin’ to try qiime2 on a real dataset as a learning experience and for comparison purposes. I’m not sure the functionality is completely there for a full comparison, but I had a question in regards to the data prep stages. Is there functionality for merging contigs, i.e. SeqPrep, ea-utils, etc? I couldn’t find any mention of it in the tutorial and wasn’t sure if it’s functionality that awaits implementation or if dada2 obviates this step or does something equivalent?

The libraries were not generated using emp, rather encompass a larger region of the 16S (V3/V4). They are not demultiplexed at this stage and utilize a different scheme (dual index, 8bp barcodes).

Sorry if I’m getting a bit ahead, but was eager to get comfortable/learn the new environment and figured I would do so using in-house data.


Hi @John,
Thanks for trying it out!

At this stage we don’t have support for paired end merging. That will come with the next version of the DADA2 plugin where there will be built-in support for that. I don’t have an ETA for that yet, but I’m hoping it won’t be too long. We’re trying to solve one technical issue that is currently blocking us.

We don’t have demultiplexing support for the scheme that you’re describing, but it might not be very difficult for us to add. Is there a reference describing the protocol (including where the barcodes are in the fastq file(s))? Alternatively, if you can demultiplex outside of QIIME right now, you can import per-sample fastq files. I can provide an example of how to do that if you’re interested in trying it out.

Thanks again!

Hi @gregcaporaso,

Sounds good. I’m familiar with merging and demultiplexing offline so not a problem…just figured I’d give it a shot in qiime2 if it was an option. I’ll ping you about the barcode scheme. You are likely at least vaguely familiar with it as it is based on the UT approach that we use at TGen North.