Why not extend the Newick format in a way that it allows both for comments
in square brackets *and* for internal labels (the ink has not yet dried..)?
It seems to me that there are now quite a number of programs that are able
to use or produce internal labels (CLUSTAL W, TreeView, TREETOOL, NJPlot,
Zharkikhs program). Obviously, these labels have somehow become one of the
standard "flavours" that is impossible to ignore. On the other hand, as
Des Higgins pointed out, it is necessary to keep all the treefiles exchangeable,
so why not keep *all* the good ideas: branch lengths after :, comments in [],
and bootstraps as node labels directly after the ).
For my own program PUZZLE (version 2.5) I now have decided to go for the
CLUSTAL W option because the real problem is, of course, to eventually
get your trees printed out!
Concerning my second proposal (about whether or not the treefile should state
in the first line the number of trees) I have received a nice idea by Frank Wright:
> I do think having a number at the top of a stack of tree
> files is not a good idea. I think the software
> should count the number of ";" in a file and thus determine
> the number of trees, or the user should be prompted for the
> number of trees in the input file. This would avoid the need
> for editing (yuk!).
I absolutely aggree! Finally, the ; 's get a function :-))
So, I change my mind and propose with Frank that *all* treefiles (intree too)
should not need to contain a specifying number!
Korbinian Strimmer
