reroot

[rcunha]

Hi everyone,

I need to Reroot the tree I've got from TreeAnnotator, because the
outgroups are not the same from my user tree.
I've tried the new version of FigTree v1.1.2 which allows rerooting
but when I do it the bars don't reajust to the new topology. Is it
possible to do, or how can I force previously the outgroup before
running the analysis?


cheers,

Regina

[Andrew Rambaut]

Dear Regina,

You shouldn't use the rerooting option in FigTree for BEAST trees as
the position of the root is something that BEAST has estimated and
all the other information is particular to this rooting position. It
may be that the tree you obtained from TreeAnnotator just happened not
to have the outgroup where you expect it (remember, TreeAnnotator
doesn't create a consensus tree, it simply picks a single tree from
the sample). You could try using tree annotator to annotate your 'user
tree' (you specify it as the 'target tree' in TreeAnnotator). This
would then give you the posterior probability of your expected
outgroup root position.

If you are certain that your outgroup is an outgroup you need to
constrain the outgroup (you can do this in BEAUti by creating a Taxon
Set that contains all you ingroup sequences and clicking the enforce
monophyly check box). Alternatively, if the outgroup is particularly
divergent from the taxa of interest, you may be better off removing
the outgroup as BEAST doesn't need it to root the tree.

Andrew

___________________________________________________________________
Andrew Rambaut
Institute of Evolutionary Biology University of Edinburgh
Ashworth Laboratories Edinburgh EH9 3JT
EMAIL - a.ra...@ed.ac.uk TEL - +44 131 6508624

[rcunha]

Hi again,

I've followed your instructions about TreeAnnotator specifying the
target tree as my user tree and it didn't work.
It gave me an error message saying Exception: null, after reading
trees.

By the way, my user tree was a NJ constructed in PAUP and saved in
PHYLIP format

Now I'm running again the analysis constraining the ingroup to be
monophyletic, the problem is that each run takes 1 and half months!!!
If there is another way....


cheers

Regina