I've been using Phylip 3.3, 3.4, and now 3.5 on a Mac (and 3.4 on Iris)
for some years now. I ran the DNAdist and NEIGHBOR programs of 3.5fat on
an old data set and was somewhat surprised that the resulting tree was
significantly different from the trees I received on all other versions of
Phylip on various platforms (which agreed with one another). Phylip
3.5fat placed two taxa which are practically identical at opposite ends of
the tree, separated by taxa which are significantly divergent from the two
in question. I've run this several times now, invoking as many options as
I thought would matter (ie jumble input order), and now change. The
3.5fat tree still completely disagrees with the other trees. Has anyone
else run into similar problems?
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Robert Rumpf "Do I have to tell a story/of a thousand
rumpf.1 at osu.edu rainy days since we first met/it's a big
Ohio State University enough umbrella/but it's always me that
Dept. of Molecular Genetics ends up getting wet" -- Sting
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
