Meta-analysis of multiple studies with different 16S rRNA gene regions using SEPP tree

Community Plugin Support
1

Hello,

I'm conducting a meta-analysis of multiple studies that target different 16S rRNA regions (V3-V4, V4). I’ve read the article (https://journals.asm.org/doi/10.1128/msystems.00021-18) and would like to construct a phylogenetic tree using the SEPP plugin for fragment insertion. However, I couldn’t find any available code specifically for building a SEPP tree in the context of a meta-analysis.

Here is my question:
After performing separate DADA2 analyses for each sequencing region, do I need to merge the resulting representative sequences again to use as input for the SEPP tree??

I would greatly appreciate it if someone could assist me with this.

2

Hello SEOK,

Welcome to the forums! :qiime2:

Here is the documentation for the Qiime2 SEPP plugin: sepp: Insert fragment sequences using SEPP into reference phylogenies. — QIIME 2 2024.5.0 documentation

It's up to you!

SEPP needs this as input:
--i-representative-sequences ARTIFACT FeatureData[Sequence]

So you could merge your DADA2 sequences artifacts and run SEPP once
or
you could run SEPP multiple times for the multiple rep-set-seqs.qza files you want to insert into the tree.

I'm not sure which is best.
(but they will both be slow :snail: )

1 Like
3

@colinbrislawn

I went with my first plan. Thanks for a quick response!

1 Like