RAxML-EPA with outgroup

[Despina]

Hello,

I am using RAxML-EPA (-f v) to place short reads in a reference tree.
The reference was built defining an outgroup (and indeed the resulting tree is correctly rooted).
Then when I use RAxML-EPA with that reference tree, the output saved in RAxML_originalLabelledTree.<NAME> doesn't keep the root.
This happens with and without specifying the outgroup in RAxML-EPA (-o option).
Is there a way to preserve the rooting based on the original tree, or take the -o option into account for RAxML-EPA?

For reference, these are the commands I'm running:
# Reference tree
raxmlHPC-PTHREADS -f a -n REFTREE -s ref_aln.fa -T 16 -x 1234 -N 100 -m GTRGAMMAI -p 1234 -o S003616303_Escherichia_coli
# EPA
raxmlHPC-PTHREADS -f v -G 0.2 -m GTRGAMMAI -n EPARUN -s ref_aln_plus_reads.fa -t RAxML_bestTree.REFTREE -T 14 -o S003616303_Escherichia_coli

Thank you,

Despina

[Alexandros Stamatakis]

Dear Despina,

I am a bit reluctant to implement this in the EPA right now.

The thing is that an outgroup simply represents a tree drawing option,
that is, you transform the still unrooted tree that includes the
outgroup by putting the root onto the branch that leads to the outgroup.

As such, using almost any tree drawing program you can just do the
rooting a posteriori by placing the a root onto the branch leading to
the outgroup.

We will be able to provide more help if you tell us how you intend to
post-analyze your evolutionary placements.

Alexis

> --
> You received this message because you are subscribed to the Google
> Groups "raxml" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to raxml+un...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>

--
Alexandros (Alexis) Stamatakis

Research Group Leader, Heidelberg Institute for Theoretical Studies
Full Professor, Dept. of Informatics, Karlsruhe Institute of Technology
Adjunct Professor, Dept. of Ecology and Evolutionary Biology, University
of Arizona at Tucson

www.exelixis-lab.org

[Despina]

Dear Alexi,

I am using EPA to place shotgun illumina 16S SSU reads into a reference tree.

The reason I would like to have the output tree of EPA rooted, is because I want to see the placement results at different taxonomic levels.

Eg, I m using a script to collapse nodes based on a predefined input, and at the same time summarize all the reads mapping to children nodes

(instead of starting over the analysis with a new reference tree that only represent the levels of interest).

For this the rooting is critical. For now, I am using a tree reading software to reroot the EPA tree and then use my script.

I was just wondering if I could directly use the EPA for that reason, and avoid that manual step.

Thank you, 

Despina

[Alexandros Stamatakis]

Dear Despina,

> I am using EPA to place shotgun illumina 16S SSU reads into a
> reference tree.
>
> The reason I would like to have the output tree of EPA rooted, is
> because I want to see the placement results at different taxonomic
> levels.
> Eg, I m using a script to collapse nodes based on a predefined input,
> and at the same time summarize all the reads mapping to children nodes
> (instead of starting over the analysis with a new reference tree that
> only represent the levels of interest).
> For this the rooting is critical. For now, I am using a tree reading
> software to reroot the EPA tree and then use my script.

I see, that makes sense.

> I was just wondering if I could directly use the EPA for that reason,
> and avoid that manual step.

I'll try to add this option to RAxML, but it may take some time, since I
already have a huge backlog of feature requests.

Alexis

[Alexandros Stamatakis]

Dear Despina,

Sorry that it took me so long. In fact there is an output format that
maintains the rooting of the tree. The EPA algorithm will produce a file
called .jplace I think according to the format specification here:

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0031009

In there you should find a rooted tree.

Let me know if that works for you,

Alexis