New thread: The Felsenstein-Zone

[James Francis Lyons-Weiler]

On Wed, 10 Dec 1997, Thomas Schlemmermeyer wrote:

> I want to come in with a completely new thread. In one of the previous
> messages there was an exotic area described, called Felsenstein-zone.
> Very bad things are supposed to happen there.

        Indeed.
>
> Well, as I already outlined in one of my previous messages I'm just
> a beginner who tries to understand it all.
> So far, my autodidactic efforts resulted in the following observations:
> 1.) many people prefer HENNIG86 to PHYLIP.

        They are used to ask different questions that are related in
        contorted ways.

.
> 2.) James Carpenter does not like Felsenstein-software at all and recommends
>     NONA, an argentinan program.

        And???

> 3.) A very intelligent, old systematist (Padre Moure) presented an oral
>    comunication at the Ribeirao Preto-Bee Congress in 1996, saying that
>    actually only the Felsenstein-package was able to correctly recognize
>    the phylogenetic groupings he constructed artificially.

        Others have shown cases where NJ outperforms others in this
        respect as well.
.
> 4.) Thiago, an intelligent student, told me, that this was purpose. Padre
>    Moure already did know the weaknesses of the programs before and
>    constructed with purpose taxons which could only be recognized by the
>    Felsenstein-programs.

        One answer is that the result communicates the different
        capabilities of distinct methods.
.
> 5.) Last not least: Felsenstein's program package is free, the others not.
>
> OK, anyone want to comment on the Felsenstein-zone, so that I may get a more
> real framework to deal with these questions.


        I and another (Hoelzer) have a paper due out this month (?)
        in Molecular Phylogenetics and Evolution that demonstrates
        the utility of a new graphical diagnostic for the detection
        and identification of long edge taxa.  Until the paper
        comes out, interested parties can read the manual to RASA 2.1,
        my software that implements this and other new tree-free
        tests.  Like PHYLIP, it's free, and can be downloaded at

        http://loco.biology.unr.edu/archives/rasa/rasa.html

        The software is for the Mac.  Mac users who don't have decent
        access to the web can contact me and I'll email them the
        software as an attached file.

        James Lyons-Weiler