Best phenetics software?

[Nicholas S. Lander]

Murray Dawson wrote:

>Would anyone like to contribute their thoughts on the "best" software for
numerical taxonomy?

>Criteria of a prospective package: user-friendly, flexible, powerful
(universal requirements!); relevant to taxonomists and well-suited for
handling both qualitative and quantitative characters; capable of cluster
analysis and ordination.

>PATN (Australian-developed) and NTSYS are two I have heard of - any
others?

Although not widely advertised, the DELTA system for descriptive taxonomy
includes software for phenetics analysis, called PCLASS. The following is
taken from the documentation CLASS.DOC

"The following programs work on distance matrices produced by the program
DIST. They are currently undocumented. To run them, enter the program name
and respond to the prompts, which request information such as the number of
OTU's (taxa) and the name of the distance-matrix file.

"PCLASS  Agglomerative clustering by combinatorial sorting strategies.
"NSIM    Lists nearest neighbours of taxa.

"Trees produced by PCLASS can be plotted on PostScript printers, as
follows. PCLASS produces a fusion table on a `data-transfer' file with
extension .TRE (e.g. GRASS.TRE). Run TXPLOT, and enter the name of this
file when the name of the `input file' is requested. Output is produced on
a file with extension .PLT (e.g. GRASS.PLT). This file is converted to
PostScript and printed by running PLTPRINT, e.g.PLTPRINT GRASS
PLTPRINT.BAT, which invokes the program PLT2PS.EXE, may need to be modified
to suit local conditions. As supplied, it defaults to A4 paper, and
printing is done by means of a PRINT command.

"The public-domain program Ghostscript (not supplied) can be used to
display PostScript files on VGA screens or to print them on various
non-PostScript printers."

The point is that data coded in DELTA can be passed immediately to PCLASS.
Equally well it can be passed to formats suitable for input to cladistics
programs including Hennig86, PAUP, MACLADE.  And, of course, once coded in
DELTA format one's data can be used to generate a plethora of other outputs
including natural language, printed keys, interactive keys, distance
matrices, HTML, RTF, etc.



-------------
Original Text
From: Murray Dawson <DawsonM at LANDCARE.CRI.NZ>, on 7/8/97 11:33 AM:
Greetings,

Would anyone like to contribute their thoughts on the "best" software for
numerical taxonomy?

Criteria of a prospective package: user-friendly, flexible, powerful
(universal requirements!); relevant to taxonomists and well-suited for
handling both qualitative and quantitative characters; capable of cluster
analysis and ordination.

PATN (Australian-developed) and NTSYS are two I have heard of - any
others?

PLEASE let us avoid the age-old arguments of phenetics versus
cladistics!

I have followed the recent string on the "best" herbarium DBMS software
with interest - it is really productive to gain a rough consensus based on
such a large pool of experience.

Murray Dawson