polytomies

[Rachel]

Hello-
I'm trying to post this in the FigTree group as well, but I want this
group's input too.

I work with a large dataset (>215 taxa) of viral sequences sampled
over time. I am having trouble preparing a tree that I feel is
sufficient for publication. My specific trouble is regions of the tree
that have poor posterior support. I feel that that nodes with poor
support are better represented as being collapsed into a polytomy than
a strictly enforced series of bifurcations. This is biologically
relevant for the virus I am analysing.

I'd like to clarify several points with the creators of the BEAST
suite as well as ask everyone's opinions.

TreeAnnotator- does the posterior probability limit mean that only
nodes with this value or greater will be included/drawn in the target
tree, or only that nodes with this value or greater will be annotated?
I've heard both explanations, but in practice neither happens for me-
I get nodes drawn and annotated with posterior probabilities far lower
than whatever I set in this program.

FigTree- I cannot find a method by which I can collapse nodes with
poor support back to an ancestral node with good support. As I
understand it, this is not 'pruning' (removing specific branches/taxa
from the drawn tree). When I use FigTree to export trees, none of the
options (nexus or newick formats, with or without annotations, etc)
produces a file with posterior probability data included in with the
taxa and branch length. Using programs like MEGA, Mesqute, and
Dendrogram are therefore fruitless with any of the file formats I get
from FigTree, since there are no posterior probabilities in the files
these downstream can use. Mesquite can handle the original .trees file
from BEAST, but that appears to require additional laborious (and
redundant) calculations of essentially bootstrapping of the sampled
trees for node frequencies.

So how would you all accomplish my goal of drawing polytomies?

thank you,
Rachel

[Pip Griffin]

Yes I'd like to be able to collapse the polytomies too, but it doesn't
seem to be possible in FigTree.

I did try exporting the tree and opening it in other programs (e.g.
Mega, which I know can collapse polytomies to combs) but the posterior
probability values don't get transferred even if I use the 'Save as
Currently Displayed' option when saving the tree in FigTree.

So in the end I've been doing it by hand in Illustrator after
exporting the tree image - pretty slow and not ideal. Maybe the
developers could include this option in the next version...

Pip

> --
> You received this message because you are subscribed to the Google Groups "beast-users" group.
> To post to this group, send email to beast...@googlegroups.com.
> To unsubscribe from this group, send email to beast-users...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/beast-users?hl=en.
>
>

[DCB]

You should also be able to accomplish some of your goals of
manipulating trees (i.e., collapsing nodes), plotting support values,
etc. using DendroPy: http://packages.python.org/DendroPy/.

See the DendroPy tutoria for details: http://packages.python.org/DendroPy/tutorial/index.html

Some basic pruning and manipulation can also be done in R using the
variety of packages available. You can download the R task view for
"Phylogenetics" and then check out some of the functions contained
within this array of packages: http://cran.r-project.org/web/views/

D

--
David C. Blackburn, Ph.D.
Division of Herpetology
Natural History Museum and Biodiversity Institute
University of Kansas

Email: david.c....@gmail.com
Web: http://www.people.ku.edu/~blackburn

[Rachel]

Thanks for the responses, everyone. both here and sent to me
privately.