Kent E. Holsinger HOLSINGE at uconnvm.bitnet
Mon Nov 26 08:24:43 EST 1990


I'd like to second Richard Thomas' suggestion that the next version of
PHYLIP concentrate on what are already its strengths and not worry too much
about a nice graphical interface.  The single most useful thing about PHYLIP
to me is that it can be used with very little effort on a wide variety of
machine architectures.  Bob Jansen and I may be the only people in the world
who still do some of our parsimony analyses (at least the bootstrapped ones)
using PHYLIP.  Even PAUP 3.0k running on a MacIIfx just isn't fast enough to
get us results from some of our analyses in less than a week.  With PHYLIP
on the UConn mainframe many analyses are possible that couldn't be attempted
otherwise.  It's a safe bet that MVS is not going to have an X Windows
implementation any time in the forseeable future.

I have only one specific suggestion, but it involves changes to one of your
parsimony programs, either MIX or DOLLO (I'm not sure where the change would
be most easily incorporated).  I'd like to see a stepmatrix implementation
similar to the one in PAUP.  This would allow parsimony analyses in which
gain/loss of restriction sites and transition/transversion substitutions are
differentially weighted.

-- Kent

More information about the Mol-evol mailing list