More on the 'cladistics' of sequences

[B.J.Tindall]

David
Some quick answers:

At 10:55 11.6.2004 +0100, David Williams wrote:
>I cannot claim to have read all the posts on this subject but it has been 
>with increasing frustration. Reciting what might be known, or assumed, or 
>understood has no ultimate bearing on the problem outlined by John Grehan. 
>In short, he has a very good point. What, if anything, are the entries in a 
>matrix?

The entries if the matrix are data (phenotypic or genotypic), unless you
have treated them in some fashion, such as creating a similarity
matrix...... Thus a set of aligned sequences is also a matrix displaying
the character state of the bases at homologous positions (we hope), and the
dervied similarity matrix a conversion of the original data heading towards
phenetic evaluation. 

>
>Most (all?) of the posts recite issues concerning methodologies, as if the 
>solution can ultimately be found in some form of numerical analysis. 
>Perhaps this is what leads Pierre, for example, to conclude that “Some 
>‘cladists’ effectively rediscovered phenetics (Jaccard index). They belong 
>to some trends of the ‘pattern cladistics’ school, which is not 
>evolutionary, thus you shouldn't bother if your own goal is phylogeny 
>inference”, an assertion based on no evidence or published source, as far 
>as I am aware, and entirely of his own making. Efforts to substantiate one 
>methodology over another is ultimately pointless. Thus, for example, to 
>equate cladistics with parsimony methods is false.

In an earlier posting I implied that some form of parismony is implicit in
most of the methods. As for "pattern cladistics" - as an outsider to the
field I, perhaps, came to the mistaken conclusion that "true cladistics"
first of all weighted characters and then one constructed the tree - I had
the feeling that pattern cladistics tended not to do that, which brings me
back to the e-mail from David Orlowich, which seems (correct me if I am
wrong) that parsiomony analysis of sequences tends to look a bit more like
pattern cladistics..... Still I suppose someone from the HNM should know
the problems caused by pattern cladistics under that roof some 20 years
ago. I agree that we seem to be going around in circles here.


>
>Once we recognise that Cladistics is about relationships, the real 
>Cladistic revolution was the reform of palaeontology and that all numerical 
>methods are all inherently phenetic (because of the matrix and thus not 
>phylogenetic) will we be able to move on.

Can't agree with the evaluation of "all numerical methods are all
inherently phenetic". If cladistics is about character weighting then there
is no reason why you can't weight characters and then let a PC do the rest
- depends at what level the "numerical analysis" is introduced. 

>
>So
 What, if anything, are the entries in a matrix?

Data, which are neither phenetic nor cladistic, nor phylogenetic, but which
may be used in different forms of analysis, including the discovery of the
"evolutionary history of the organisms concerned".
Brian

>
>David.
>


********************************************************************
* 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                                            *
*                                                                  *
* Homepage: http://www.dsmz.de/index.html                          *
* E-MAIL: contact at dsmz.de (general enquiries)                      *
*         sales at dsmz.de (sales)                                    *
********************************************************************