Qiime1 to qiime2 transition paired end as not-barcoded

I am currently transitioning from qiime1 to qiime2.
I have some doubts on the best way to treat Illumina paired end data (we use 2x300 run on Miseq) using Illumina protocol.
In the past I did the following (I mention only relevant, not quality checks)

  • joining paired ends reads
  • use the qiime script split_libraries_fastq.py as in the following (using a make script) treating samples as non barcoded.
    Is there any suggestion on doing the same for qiime2? Or other better practice to pass to qiime2 analysis in that condition?

slout_single_sample_q20/seqs.fna: (addprefix (RAW_DATA_DIR_JOINED)/,(addsuffix .fastq.gz, (SAMPLEJOINED)))
condactivate qiime1 ;
-i $$(perl -lpe ‘s/\s+/,/g’ <<<"$^")
–sample_ids $$(perl -lpe ‘s/_/./g; s/\s+/,/g;’ <<<"$(SAMPLEJOINED)")
-o slout_single_sample_q20/ -q 19 --barcode_type ‘not-barcoded’ ;\

HI @MichelaRiba,

As you transition, you might want to look into the QIIME 2 for experienced microbiome reserachers tutorial, which does some comparisons between QIIME 1 and QIIME 2.

It sounds like you’re dealing with already demultiplexed data (a more and more common format). I’d recommend looking at the manifest format for fastq importing to see if it fits your needs. You may be tempted to use the casava import format, but a lot more people seem t have success with the manifest version.

The second issue as you switch from QIIME 1 (which was updated c. 2015) to QIIME 2 would also be to update from OTUs to denoising. I mention this because it affects how your reads are joined. If you choose to use DADA2, it will wrap the quality filtering, denoising, and read joining steps into a single command. This may take a while to run). There’s an example of Dada2 for paired end data in the Atacama Soils tutorial. Alternatively, if you don’t mind a few additional steps and want something a little faster with a consistent run profile, the Alternative Methods of Read Joining tutorial goes over how to run deblur on paired end sequences.



Hi, In the end I have been successful in data import using pre-joined sequences:
qiime tools import
–type ‘SampleData[JoinedSequencesWithQuality]’
–input-path manifest.csv
–input-format SingleEndFastqManifestPhred33
–output-path joined.qza

I treated sequences as single read, forward

best regards,



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