bootstrap values

John Harshman harshman.diespamdie at sjm.infi.net
Thu Mar 22 20:58:10 EST 2001


In article <99dden$sle$1 at mercury.hgmp.mrc.ac.uk>, Kerstin Hoef-Emden
<kerstin.hoef-emden at uni-koeln.de> wrote:

> Hi to all,
> 
> some time ago, I read a small thread, I cannot remember whether in this
> group or another related one, about where to place the bootstrap values
> in a tree. One response was, that in rooted trees the bootstrap values have
> to be placed at the nodes, but in unrooted trees on the branches. 
> 
> Does somebody know references addressing this item?

No, but I can think. On a rooted tree, there is a one-to-one
correspondence between a node and the branch directly below it, so it
doesn't matter which one you put the bootstrap value next to. On an
unrooted tree, a node is just one end of a branch, and a branch is the
line dividing the taxa into two groups. And *that* is what a bootstrap
value is properly applied to. It's simpler if you always attach the value
to a branch.

Now, there's one spot where that becomes tricky: the root. You should
realize that unless your characters are pre-polarized, the root really
doesn't exist. It's not a node. Therefore the bootstrap support for the
two lowest nodes (one on either side of the root) is just the support for
a single branch: the branch connecting the ingroup taxa to the outgroup
taxa. There is no separate bootstrap value for ingroup monophyly and for
outgroup monophyly, just one value for the existence of a branch
separating the two. And of course if you have a single outgroup OTU,
there's no bootstrap value for that branch at all, because it's just a
terminal branch.

-- 

   *Note the obvious spam-defeating modification
    to my address if you reply by email. 





More information about the Mol-evol mailing list