Tree building from post deblur files

[myor (Michael)]

Hi

I downloaded and merged a few deblurred biom tables from qiita and want to build a tree in QIIME2.
The deblur was done without hashes, while the rep-seqs comes from hashes (currently the repseqs are in the biom table). Is there a workaround to create a repseqs file?

Thanks
Michael

[antgonza (Antonio Gonzalez Pena)]

Hi @myor,

This is a similar question than:

Importing deblur data from QIITA

I'm working on a meta-analysis of datasets from QIITA for analysis in qiime2. I've been able to download datasets processed using deblur from QIITA as well as their respective metadata files. I've merged the BIOM and metadata files but in order to continue with my analysis, I require a rep-seqs file. It looks like this file isn't generated by QIITA from data processed with Deblur. Is there a way I could extract the rep-seqs file from the deblur BIOM file in order to run "qiime diversty core-metr…

However, some of the steps in Qiita have changed and I think having a shorter answer here is worth it. Anyway, the biggest difference is that we shouldn't have the lower/upper case issue not present anymore so we just need to run:

biom summarize-table --observations -i your_biom.biom \
    | tail -n +16 | awk -F ':' '{print ">"$1"\n"$1}' > rep_seqs.fna

Then import your biom and fna to QIIME2 using the regular import commands.

Hope this helps.