Join and demultiplex paired-end sequences

hi @thermokarst,

Would you please provide more details?

If I import paired end reads respectively, how can I join the reads and and demultiplex barcodes?
The Casava format in the tutorial is a demuxed one.

The following information was written at the beginning of 2017, before QIIME 2 had support for pair-end analyses. If you are interested in analyzing your pair-end data, please check out the Atacama Tutorial for a starting point! If you get stuck, feel free to open a new topic on the forum!

Outdated Information

Hi @yech1990! As @thermokarst noted, we can assist you with importing your paired-end sequence data into a QIIME 2 artifact, but you won’t be able to do anything with the artifact in the latest version of QIIME (2.0.6). We are planning to add paired-end read support to the QIIME 2 dada2 plugin for the next release (scheduled for February), so it will be possible to denoise paired-end data.

In the meantime, here’s a couple of options:

  • If you’re wanting to analyze paired-end data: I recommend using QIIME 1’s or another external program that joins reads (e.g. pandaseq, fastq-join, etc), and then demultiplex your data using After that, use one of the QIIME 1 OTU picking methods (e.g. and then import your .biom file into QIIME 2 following this tutorial. You can then perform QIIME 2 analyses using your imported feature table .qza file.

  • If you’re okay with analyzing single-end reads: I recommend using your forward (R1) reads. If your multiplexed sequences were generated using the Earth Microbiome Project protocol, you can import and demultiplex them following this tutorial. You can then use the demultiplexed sequences as input to qiime dada2 denoise to obtain a feature table. A couple of important notes if you’re using dada2 to denoise your sequence data:

    1. dada2 expects your demultiplexed sequences to come from a single sequencing run. If you have sequence data that came from more than one run, you’ll need to denoise each run separately with dada2, and then merge the results. See the FMT tutorial for an example of how to do this.

    2. dada2 denoise expects the sequence data to be single-end (i.e. not already joined). In the next QIIME 2 release we expect to support paired-end data, but for now the QIIME 2 dada2 plugin only supports single-end reads.

Get in touch if you have any questions!


An off-topic reply has been split into a new topic: Importing demultiplexed sequences in Casava format

Please keep replies on-topic in the future.