Duplicate samples in feature tables during merge

[ben (benjamin w.)]

Hey thank you for the help. I have a perhaps a closely associated question:

I have 7 MiSeq runs, they have been DADA'd and denoised:

qiime feature-table merge \

--i-tables ~/QIIME2_4_DADA2/180209_MSQ72/table-1.qza
--i-tables ~/QIIME2_4_DADA2/180215_MSQ73/table-2.qza
--i-tables ~/QIIME2_4_DADA2/180216_MSQ74/table-3.qza
--i-tables ~/QIIME2_4_DADA2/180417_MSQ77/table-4.qza
--i-tables ~/QIIME2_4_DADA2/180420_MSQ78/table-5.qza
--i-tables ~/QIIME2_4_DADA2/180509_MSQ79/table-6.qza
--i-tables ~/QIIME2_4_DADA2/180518_MSQ80/table-7.qza
--o-merged-table ~e/QIIME2_5_merge/data/merge.table.qza

I receive an error:

Plugin error from feature-table:

Same samples are present in some of the provided tables: Blank

Debug info has been saved to /var/folders/82/86pyg4s12_vb2mzdf8yknlrh0000gn/T/qiime2-q2cli-err-gci2w6y6.log

I assume it is because samples in the MiSeq runs have the same title. Which will need to be addressed. How should I go about this? Or when/where should each of these samples be renamed?

Interestingly, I was able to merge the seq.qza files with success:

qiime feature-table merge-seqs \

--i-data ~/QIIME2_4_DADA2/180209_MSQ72/rep-seqs.qza
--i-data ~/QIIME2_4_DADA2/180215_MSQ73/rep-seqs.qza
--i-data ~/QIIME2_4_DADA2/180216_MSQ74/rep-seqs.qza
--i-data ~/QIIME2_4_DADA2/180417_MSQ77/rep-seqs.qza
--i-data ~/QIIME2_4_DADA2/180420_MSQ78/rep-seqs.qza
--i-data ~/QIIME2_4_DADA2/180509_MSQ79/rep-seqs.qza
--i-data /~/QIIME2_4_DADA2/180518_MSQ80/rep-seqs.qza
--o-merged-data ~/QIIME2_5_merge/data/rep-seqs.qza
Saved FeatureData[Sequence] to: ~/QIIME2_5_merge/data/rep-seqs.qza

Thank you. Ben

EDIT:

It looks like the step I need to redo is at the:
'qiime demux emp-paired ' Step, we run blank wells on all our runs to see what the background microbiome may look like and this also helps us control for background intrusion. I may need to re-run most of this - however it looks like there are only 3 blank samples that have the same names. Could I just go head and force the merge despite this? It will take me a couple of days to redo the demux-->dada/denoise step.

[Nicholas_Bokulich]

@ben

I assume it is because samples in the MiSeq runs have the same title. Which will need to be addressed. How should I go about this? Or when/where should each of these samples be renamed?

You can rename samples in a feature table using this trick. No need to re-do anything.

Another option would be to filter out the Blank samples.

@ben

Interestingly, I was able to merge the seq.qza files with success:

Because that file just contains feature IDs and sequences, not sample IDs (which cause the issue when merging tables).