[Taxacom] Molecular vs. morphological evidence

[Paul J. Morris]

On Wed, 1 Jul 2009 08:06:40 -0400
"John Grehan" <jgrehan at sciencebuff.org> wrote:
> Should? Whole 'genomes' (read entire sequence) is still the same kind
> of data, and taking a chunk or taking the whole (and that would be
> for all species) doesn't change the ingredients and the inherent
> problems of molecular bean counting.

Wrong.

Many of the people working with whole genomes are identifying
multiple classes of molecular homologies in addition to sequence
similarity.  Read this paper and its supplemental material:

Sark A., et al., 2007. Discovery of functional
elements in 12 Drosophila genomes using evolutionary signatures.
Nature 450:203-218.

Among other things, based on patterns of which bases are changing
amongst Drosophila species, they predicted the presence of previously
unknown protein coding regions, identified regions of both
pre-translational and post-translational regulation, and constructed a
tentative map of regulatory interactions between several thousand
genes.  The assertion that region X of the genomes of twelve fruit fly
species is a protein coding exon, and that the protein gene product of
this region regulates the production of another protein in region Y
involves multiple claims of homology.  That is, homology in the
classical senses of Saint-Hilaire and Richard Owen, though the
language they are using to describe these are "functional elements".
Each part of that assertion is inherently different from, and has
different inherent problems from the molecular bean counting of
sequence similarity.  

This is one example of many.  Phylogenomics is a different world.

-Paul
-- 
Paul J. Morris
Biodiversity Informatics Manager
Harvard University Herbaria/Museum of Comparative Zoölogy
mole at morris.net  AA3SD  PGP public key available