[Molecular-evolution] Re: Phylip

Mary K. Kuhner mkkuhner at kingman.gs.washington.edu
Tue May 2 16:27:05 EST 2006


In article <mailman.799.1146600165.16885.mol-evol at net.bio.net>,
Antonis Chatzinotas  <antonis.chatzinotas at ufz.de> wrote:

[please don't post in HTML, it's hard to reply!]

>I have used Phylip to make a parsimony tree (Seqboot (100
>replicates) &gt; dnapars &gt; consense). When I draw the consensus tree
>in Treeview I obtain a tree with a scale bar with a value
>&quot;10&quot;.&nbsp; Does this scale bar have any meaning and what
>exactly does it mean? <br>

Phylip's consensus tree routines produce a tree whose "branch
lengths" are the proportion of times that clade appeared in the
set of trees.  Since these are really not branch lengths, it is
better not to display them as such.  The standard approach is to
show either no branch lengths, or the branch lengths of the best
tree found, and to put the bootstrap numbers next to their
corresponding nodes.

The scale bar is telling you how many centimeters on your page
correspond to a clade that appears in such-and-such percent of the
trees, but you probably don't want to use that information--
better to look at the actual numbers in the tree file.

Mary Kuhner mkkuhner at gs.washington.edu



More information about the Mol-evol mailing list