Poor Willi Hennig?

Miriam Richards mrichard at FRED.FHCRC.ORG
Sun Sep 17 12:01:53 EST 1995


I am one of "those molecular people" who finds these occasional flames 
about molecular phylogenetics rather peculiar.  I sometimes think that 
the extreme points of view such as were espoused by Ludvig Mortberg 
at the beginning of this thread, reflect the view that systematists 
using molecular data are upstarts who never bothered to learn the true 
and difficult methods of morphological systematics, and therefore don't 
have any understanding of how organisms really function or evolved.  In 
fact, how organisms evolve is not the only interesting question to be 
addressed with molecular data.  Some of "those molecular people" are also 
interested in how genes evolve.  Ludvig Mortberg's statement illustrates 
that this has not occurred to him - he assumes that all molecular 
systematics is aimed at elucidating the relationships among organisms:

Mortberg wrote:
> I think that sequence information should only  be used if nothing else
> works. If you can't identify any reliable morphological traits try
> sequencing a couple of genes with broad outgroups. What to sequence is
> hard to decide. Go for something linked to morphology or metabolism.
 
Currently, I am interested in the relationships among certain genes of 
the immunoglobulin superfamily.  How would finding something linked to 
morphology or metabolism help me with this problem?  Plainly it would 
not.  More importantly, how do I choose an appropriate outgroup for an 
ancient family of vertebrate genes?  I cannot simply take the sequence 
for sharks or protochordates (if the sequences existed) as outgroups even 
though that would be appropriate if I were trying to understand vertebrate 
phylogeny. I have to settle for genes that appear to primordial based on 
the structure of their protein products, and align them according to 
similarity of both sequence and structure.  While I would love to be sure 
that I know which characters are homologous, similarity is the best I can 
do at the moment.  Perhaps die-hard anti-molecular types would suggest 
that I should not even try such analyses, but that point of view that 
would be completely antithetical to the spirit of scientific inquiry.

Just for the record, I am one of "those molecular people" who uses both
parsimony and outgroup analysis whenever possible, having been trained by 
a tough morphological systematist who also uses molecular data.  But not 
being a religious person, I have occasionally also been known to do 
neighbour-joining and maximum likelihood analyses.


Miriam Richards
Fred Hutchinson Cancer Research Center
Seattle, Washington




More information about the Mol-evol mailing list