joe@GENETICS.WASHINGTON.EDU (Joe Felsenstein) (06/06/90)
My apologies for posting a 153-line message. The following is material
from the PHYLIP version 3.3 manual about programs distributed by others.
Since I have just used the net to advertise my own programs, the programs
by others should be mentioned too.
Here are some of the other phylogeny packages that I know about:
1. David Swofford (Illinois Natural History Survey, Natural Resources
Building, 607 East Peabody Drive, Champaign, Illinois 61820, U.S.A.)
distributes PAUP (Phylogenetic Analysis Using Parsimony).
Since December, 1985, Swofford has been distributing a precompiled
executable object-code versions of PAUP for the IBM PC and other MSDOS systems.
As of this writing (May, 1990) he was about to release version 3 (PAUP/Mac) for
the Macintosh, and later in the year hopes to release version 3 for PCDOS
systems and ultimately for mainframes. Its cost is $50. Orders received for
the Mac version before its release will be filled but the printed documentation
will arrive later, after the formal release.
PAUP 3.0 is probably the most sophisticated parsimony program. It allows
multistate characters, user-defined weights on individual state transitions,
Wagner, Camin-Sokal and Dollo parsimony methods, bootstrap confidence
intervals, and finding all most parsimonious trees by branch-and-bound. It
also has provision for computing Lake's linear phylogenetic invariants.
PAUP is (a great) many times faster than the parsimony programs in PHYLIP.
2. Swofford also distributes an older package of programs, BIOSYS-1,
including some phylogeny estimation programs, for use with gene frequency data,
with particular attention to distance methods. BIOSYS-1 is distributed on an
IBM PC-formatted floppy disk. Included are precompiled versions for the IBM PC
and source code for uploading to IBM, VAX/VMS, Unix, Prime and CDC mainframes
and minicomputers. The price is $35.00. BIOSYS-2 is under development, but it
is too early to anticipate a completion date.
3. If you have a Macintosh computer and any interest in discrete-state
parsimony methods (including DNA and protein parsimony), you should definitely
get MacClade. As of this writing it was distributed by Wayne Maddison,
Department of Integrative Biology, University of California, Berkeley,
California 94720. U.S.A. David Maddison and he have produced a program
enabling you to use the mouse-window interface to specify and rearrange
phylogenies by hand, and watch the number of character steps and the
distribution of states of a given character on the tree change as you do so.
MacClade is positively addictive and will give you a much better feel for the
tree and your data. It's the closest thing to a phylogeny video game that I
have seen. It has been influential in spurring the inclusion of interaction
and graphics into other phylogeny programs. (I have tried to supply this
functionality in PHYLIP by incorporating the programs MOVE, DOLMOVE, and
DNAMOVE, which act somewhat like MacClade). MacClade does not have a
sophisticated search algorithm to find best trees: it largely relies on you to
do it by hand (which is surprisingly effective), with only a local
rearrangement algorithm available to improve on that tree. Version 2.1 was
released in March 1987. Version 3, with a spread-sheet data editor, support
for polytomies, charting facilities, PHYLIP, PAUP, and HENNIG86 compatibility,
and many other new features, is coming soon. The program is originally written
in Lightspeed Pascal but is provided in pre-compiled form, and is available for
distribution cost ($6 in North America, $8 overseas, or three or four Macintosh
diskettes in barter). It requires a 512K Macintosh, Macintosh Plus, or larger.
In the near future the distribution of version 3 will be on a commercial basis
by Sinauer Associates of Sunderland, Massachusetts, although versions 1 and 2
will continue to be freely distributable. Version 2 is also available by
electronic mail from the EMBL and Houston net servers which distribute free
software for molecular data analysis, in their Mac software.
4. J. S. Farris (41 Admiral Street, Port Jefferson Station, New York
11776, U.S.A.) has produced Hennig86, a fast parsimony program including
branch- and-bound search for most parsimonious trees and interactuve tree
rearrangement. Although complete benchmarks have not been published it may be
a bit faster than Swofford's PAUP; both are a great many times faster than the
parsimony programs in PHYLIP. The program is distributed in executable object
code only and costs $50.
5. Jotun Hein, currently at the Universite de Montreal has produced
TreeAlign, a multiple sequence alignment program that builds trees as it aligns
DNA or protein sequences. It uses a combination of distance matrix and
approximate parsimony methods. The program is large, and is available in C
source from either the EMBL or the Houston net servers, from which you can
fetch PC, Mac, or VAX programs free for molecular data. On EARN or BITNET you
can send an electronic mail message to NETSERV@EMBL with a "Subject: " line
that reads HELP SOFTWARE. On BITNET or Internet you can send to the Houston
server which is: genbank-server@uhnix2.uh.edu a message that has a "Subject: "
line: SEND UNIX HELP or SEND VMS HELP to receive a file telling you how to get
further sofware from them, including TreeAlign. TreeAlign uses so much memory
that it cannot be run on PC's (DOS or Mac systems) but is really designed for a
workstation or mainframe. You can also get TreeAlign over Internet by
anonymous ftp from lavaca.uh.edu from the appropriate subdirectory of
pub/genbank-server.
6. David Penny (Department of Botany and Zoology, Massey University,
Palmerston North, New Zealand) has offered (I believe for free) a fast
parsimony program, TurboTree. Contact him for details.
7. Andrey A. Zharkikh, Yu. V. Rzhetsky, and their co-workers in the
Institute of Cytology and Genetics, Siberian Branch of the Academy of Sciences
of the USSR, Novosibirsk 630090, USSR are offering VOSTORG, a package of
programs for alignment and analyzing phylogenies for molecular sequences on IBM
PC-compatibles that include some rather fancy graphics. I do not know the
exact conditions of distribution: Zharkikh and Rzhetsky ask that interested
parties contact them at that address.
8. Joyce Miller of the Department of Plant Pathology, University of
Wisconsin, 1630 Linden Drive, Madison, Wisconsin 53706 is offering RESTSITE, a
program for PC-compatibles to construct phylogenies from restriction site or
restriction fragment data. It is distributed for free (send her two 360K PCDOS
floppy diskettes) and includes programs to compute genetic distances from
restriction site or fragment data and two PCDOS computer programs to do UPGMA
and neighbor-joining methods for constructing phylogenies from the distance
matrices. Some source code is also included.
9. J. S. Farris earlier released a package of phylogeny programs, PHYSYS,
which, at about $5,000, was extremely expensive (in my opinion, which is
certainly a biased one).
10. Kent Fiala's program CLINCH is available for doing compatibility
analysis of characters. Fiala's program finds largest cliques of characters.
It has more options and can handle more kinds of data than the compatibility
program in PHYLIP. It is available in both mainframe and IBM PC versions (PC-
CLINCH), for free. It is written in FORTRAN, and for the IBM PC the compiled
source code is also distributed. For information, write Kent Fiala (Department
of Ecology and Evolution, SUNY, Stony Brook, New York 11794, U.S.A.),
11. Walter Fitch (Department of Ecology and Evolutionary Biology,
University of California, Irvine, California 92717, U.S.A.) has available, as
far as I know for free (or for media and mailing cost), a large number of
FORTRAN programs for not only estimating trees but doing various other
manipulations of data that might be needed in studies of molecular evolution
(for example, looking for homologies among sequences). These programs were
originally written for the Univac system at Wisconsin He has gradually been
rewriting these to work on microcomputers, and as he does so, Fitch will be
releasing them; some are available already.
12. Philip Taylor (MRC Institute of Virology, University of Glasgow,
Glasgow G11 5JR, U.K.) offers a FORTRAN-66 program, DENDRON, for distance
matrix analyses, which he says obtains results superior to those of my program
FITCH (it does not, comparing his examples with my current version of that
program). The conditions of distribution are not stated (Taylor, 1988).
13. Christopher Meacham (University Herbarium, University of California,
Berkeley, California 94720, U.S.A.) distributes COMPROB, a FORTRAN program to
compute probabilities that characters would be compatible at random, thus
telling us which clique is "most surprising". Chris also makes his phylogeny
plotting programs PLOTREE and PLOTGRAM available. These were formerly
distributed in PHYLIP (up to version 3.2). Some code from them is incorporated
into DRAWGRAM and DRAWTREE, but particularly PLOTREE is different in the kinds
of diagrams it can draw from either of those programs.
-----
Joe Felsenstein, Dept. of Genetics, Univ. of Washington, Seattle, WA 98195
Internet/ARPANet: joe@genetics.washington.edu (IP No. 128.208.128.1)
BITNET/EARN: FELSENST@UWALOCKE
UUCP: ... uw-beaver!uw-entropy!uw-evolution!joe