Info req.: molecules/morphology

[B. J. Tindall]

Oops, sorry I should have looked at your address and would have seen you
were not a zoologist!!
I will only refer to work in microbiology, where there is a tendancy to
rely on the molecular data (but that is a longer story). The first problem
is to recognise is that there may be problems with getting "absolute"
answers with molecular data. There are two basic approaches which have been
used in microbiology. One is to take a particular programme, produce a
"tree" and then use methods such as boostrapping to test the significance
of the branching order (which would also link in with the significance of a
particular grouping). The method seems to fail when you get internal
branches which are relatively short compared to the longer branches. To get
around this problem the current school of thought which is evolving is that
one uses several different methods of analysis (i.e. neighbor-joining,
parsimony, and maximum likelihood). One then compiles a consensus tree,
which contains collapsed clades at points where either the various methods
cannot resolve the problem, and/or where the different methods are also not
in agreement. Unforutnately, the more dense the "tree" gets the more of a
problem this becomes, and it means that one may be able to define a major
clade, but subdivisions within the clade (perhaps what one is trying to
solve) just can't be resolved. Both of these approaches are general
principles. There are at least a couple of publications dealing with this
type of problem.
I think there must have been several articles on comparing different trees
produced by different gene sequence comparison, which is the molecular vs.
molecular approach. However, the molecular vs. non-molecular approach is
rather neglected in microbiology. I started looking at this type of
approach some years ago. Essentially one is looking at non-molecular data
to check whether a particular grouping/branching order is reflected in both
the molecular and non-molecular data, or to use non-molecular data to
resolve the unresolved groups/branching order. Using a Hennigian based
phylogenetic method is rather problematic in microbiology, so I would not
claim that I am reconstructing phylogeny. I made passing reference to this
type of approach in the proceedings of a meeting in 1993, but it did not
take off, simply because the problems with the sequence analysis which I
mentioned above were not apparent. However, I still use non sequence data
to evaluate results produced by sequence analysis. Perhaps this is one
reason why I do not always agree with Ken Kinman, but differences in
opinion makes the World go around. There is a logic basis for using such
methods, but one very quickly ends up in the philosophical swamp, dealing
with problems which have been around for a number of years, which include
which data sets to compare and why etc.
Perhaps this helps?
Brian


At 08:40 AM 6/27/00 +0200, B. J. Tindall wrote:

>Is this question just related to zoology, or do you want to diverge into
>other groups? There are quite a few hidden problems in the question, which
>include whether molecular trees are reconstructed/estimated by similarity
>matrix, parsimony, or maximum likelihood methods (all can give rather
>different topologies and of course within any one class of programs each
>algorithm many also give slightly different results).

I'm working with plants, but am thinking in terms of general
principles.  What I am seeking is some discussion of situations,
hypothetical or otherwise, in which molecular based analyses reveal clades
that are difficult(?impossible) to define/diagnose with non-molecular
characters.








Thomas G. Lammers, Ph.D.

Assistant Professor and Curator of the Herbarium (OSH)
Department of Biology and Microbiology
University of Wisconsin Oshkosh
Oshkosh, Wisconsin 54901-8640 USA

e-mail:       lammers at uwosh.edu
phone:      920-424-7085
fax:           920-424-1101

Plant systematics; classification, nomenclature, evolution, and
biogeography of the Campanulaceae s. lat.
-----------------------------------------------------------
"Today's mighty oak is yesterday's nut that stood his ground."
                                                 -- Anonymous



********************************************************************
* Dr.B.J.Tindall      E-MAIL bti at dsmz.de                           *
* DSMZ-Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH *
* Mascheroder Weg 1b, D-38124 Braunschweig, Germany                *
* Tel.: ++ 531 2616 0 (general)                                    *
* Tel.: ++ 531 2616 224 (direct)                                   *
* Fax:  ++ 531 2616 418                                            *
* Fax:  ++ 531 2616 491 (ISDN)                                     *
*                                                                  *
* Homepage: http://www.dsmz.de/index.html                          *
* E-MAIL: help at dsmz.de (general enquiries)                         *
*         sales at dsmz.de (sales)                                    *
********************************************************************