Get representative sequences from pre-joined data

Hello

maybe I’m missing something obvious, but how to I obtain representative-sequences and table from pre-joined data?

My first qiime command is like this (preprocessed means prejoined by non-qiime tools):
qiime tools import --type 'SampleData[JoinedSequencesWithQuality]' --input-path 16s-preprocessed.manifest --output-path 16s-preprocessed/16s-preprocessed.qza --input-format SingleEndFastqManifestPhred33

I want to call phylogeny (and later diversity), but that requires representative-sequences

Desired commands I want to run:

  • qiime phylogeny align-to-tree-mafft-fasttree
  • qiime diversity core-metrics-phylogenetic

I tried to do it by deblur plugin, but it failed:
qiime deblur denoise-16S
–i-demultiplexed-seqs 16s-preprocessed/16s-preprocessed.qza
–p-trim-length -1
–o-representative-sequences 16s-preprocessed/16s-preprocessed-rep-seqs.qza
–o-table 16s-preprocessed/16s-preprocessed-table.qza
–p-sample-stats
–o-stats 16s-preprocessed/16s-preprocessed-stats.qza

Plugin error from deblur:

Command '['deblur', 'workflow', '--seqs-fp', '/tmp/qiime2-archive-_hmmc69y/0c43de2e-4945-4bc2-b29a-e0cdefaccd26/data', '--output-dir', '/tmp/tmp155vizhn', '--mean-error', '0.005', '--indel-prob', '0.01', '--indel-max', '3', '--trim-length', '-1', '--min-reads', '10', '--min-size', '2', '--jobs-to-start', '1', '-w', '--keep-tmp-files']' returned non-zero exit status 1

Neither this tutorial helped:
https://docs.qiime2.org/2018.8/tutorials/read-joining/#importing-pre-joined-reads

Any idea, please?

Michal Lichvar

I found out, that I must specify --p-trim-length, otherwise deblur fails … but I do not want to trim :slight_smile: as I lost some data … is there another way to obtain representative-sequences?

Dear @michal.lichvar,

The Deblur algorithm requires all sequences are the same length – a may be needed to process your data.

Best,
Daniel

Hi @michal.lichvar,
If you want to use one of the denoising methods (Dada2 or Deblur) you’ll either have to have import non-joined reads (for Dada2) or trim them to the same length (for Deblur). Outside of these, the only other method available for you in Qiime2 would be to use vsearch to dereplicate or do OTU picking instead.

Hi Mehrbod

thank you, vsearch looks like tool I was asking for :wink:

M.Lichvar

1 Like

An off-topic reply has been split into a new topic: Can DADA2 be carried on pre-joined reads

Please keep replies on-topic in the future.

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