DADA2 generated rep.seqs in R to QIIME2 for making use of the "qiime fragment-insertion sepp"

General Discussion
1

I need advice on SEPP. Is it possible to import DADA2-generated (R) rep.seqs to QIIME2 and make fragment-insertion sepp tree?

I found this information useful, but I am not sure if using Silva 128 SEPP Reference Database would be problematic, given that I have used Silva 138 for the taxonomy.

In case it is OK to use Silva 128, would the following workflow be correct?

qiime tools import \
--input-path repseqs.fasta \
--output-path repseqs.qza \
--type "FeatureData[Sequence]

Followed by:

qiime fragment-insertion sepp \
--i-representative-sequences ARTIFACT FeatureData[Sequence] \
--i-reference-database ARTIFACT SeppReferenceDatabase \
--o-tree ARTIFACT  Phylogeny[Rooted] \
--o-placements ARTIFACT Placements

Thank you,

2

Hello Irshad,

I'm not an expert on SEPP, but I have used one database for taxonomy and another for tree-building and it seems to work okay.

The idea is that taxonomy is primarily about feature names, say for bar plots and ANCOM BC, and trees are about phylogeny, say for UniFrac distances, so they don't have to match.

You commands look good! Try it and see how it goes.