[Taxacom] Molecular data and synapomorphies

[John Grehan]

> -----Original Message-----
> From: Alan DAvid Forrest [mailto:aforrest at rjb.csic.es]


> I personally stick to simple text or fasta alignments that can be used
> virtually anywhere else without conversion. 

As an ignorant non-molecular biologist I had never heard of this format
until recently. Then I was not able to open some such files, even with
software that was supposed to do that. I had to have a colleague convert
to nexus and also direct me to some web programs before I could do
anything. I came across a website that was designed as a repository for
molecular data that would make access automatic, but not many were using
at such.

One key requirement is access to the alignments as they were aligned by
the original authors since different alignment programs may give
different results. In at least one case the data was not in its aligned
format.

Another problem I have had is identifying EXACTLY how many sequences
comprise each gene of each taxon compared. That might be easy for a
molecular systematist to figure out, but not for me.

I am not sufficiently
> knowledgable about morphological, biogeographical or cytological data
to
> comment on how such data is currently stored, but would be happy to
> learn (suggestions welcome).

At least with morphological data, one would expect accurate descriptions
and illustrations for each taxon for a particular character, or at the
very least citations to sources with such information. Unfortunately
that seems to be wanting in many morphological studies (so a counterpart
to access to molecular data) - particularly for larger groups. All too
many systematists just make proclamations about characters and
concentrate on generating 'lots' of them as if this will somehow snow
everyone on the lack of actual documentation. I've seen this in the
human origins field where various studies get acclaim for their many
characters without any attention to the fact that most don't even have
comparative illustration of their veracity. Other characters continue to
be recycled even when they have been shown to be incorrect.

> As for 'more informed interpretation' I mean simply: how can one
decide
> if a particular type of data is 'best' or 'correct' without comparison
> to other data? 

Ok. 

John Grehan