[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: molecular DNA, phylogeny, and random notes

Original Message by Stephan Pickering Thursday, 2 January 2003 15:24 

> Comparative molecular sequence analyses are important
> tools (S.B. Hedges, S. Kumar, et al., have published
> fascinating forays into the past).

Their calibrations tend to be horrible, and they have at least sometimes used 
just as horrible cladograms. Their paper in Science in 1998 had lizards as 
theropsids -- I mean, imagine it, lizards as theropsids, with the 
lizard-mammal divergence dated at 274 +- tens of Ma, while the bird-mammal 
divergence, with the famous fossil date of 310 Ma*, used as the sole 
calibration point --, a Placentalia like this:

+--"Edentata" (probably just Xenarthra)
`--+--Sciurognathi (some rodents, incl. rats & mice)
   `--+--Hystricognathi (other rodents, incl. the South American ones)
      `--+--Paenungulata (rest of Afrotheria not considered)

-- classical long-branch attraction all over the place, which results in the 
dating of divergences that have never happened. Not to mention the regularly 
ludicrous error margins which can approach half the date itself. Their 1997 
paper, also Science, about "mass survival" of birds across the K-T, misroots 
birds so as not to find Palaeognathae. And so on. More information, if 
someone's interested, offlist.

* The age of the oldest known certain amniotes, namely *Hylonomus*, which is 
a member of the sistergroup of Diapsida, and *Archaeothyris*, an 
ophiacodontid -- a basal theropsid, but more derived than varanopseids and 

> The problem with applicability of molecular sequencing
> to phylogenetic systematics is that, so far, recovered
> DNA from pre-Cenozoic material, while interesting,

if real,

> is not useful in answering phylogenetic questions.
> [...]
> In fact, I argue, the fossil record of Mesozoic
> dinosaurs is not, in a sense, "evidence" of
> evolutionary processes. A tyrannosaur skeleton tells
> one a magnificent organism once lived, but not much of
> the evolutionary processes of tyrannosaurs, the
> polarities of ancestor/descendent nodes of character
> traits.

A hundred skeletons from different times might do the job, though. Just might.

> A classic example of this are murine rodents. [...]
> The evidence comes not
> from sifting isolated rodent teeth at an excavation
> site, but from rigorous molecular data, using
> bea-blobin family of genes.

OK. That's what beta-globin says. I suspect other genes say the same, but has 
someone asked them? (I don't have an opinion about murine systematics, but 
relying on beta-globin alone is somehow similar to relying on teeth alone, 
isn't it?)

> Nuclear DNA and mitochondrial DNA obtained
>  from the amber-trapped dinosaur

Let's _hope_ DNA can survive that long, even in amber.

> My own "gut feeling" is that mitochondrial genome data
> is unreliable,

(Just for the record -- looks like many people, at least the zoologists at 
the university here, share this opinion. However, see Waddell et al., there 
is more than zero signal in there, even though the noise is considerable.)

> Thus, our amber-trapped dinosaur, and the
> emerging genomes of living dinosaurs, will have to be
> analysed with a combination of mitochondrial and
> nuclear DNA data sets.

Of course.