[Taxacom] molecular nonsense?

Bob Mesibov mesibov at southcom.com.au
Thu Nov 6 16:26:56 CST 2008

Hi, John.

I think the best way to understand 'this molecular stuff' is with

The kindest analogy is with the first microscopes. The optics were crap
and people didn't have a clue what they were looking at. Reports
appeared of dust mistaken for 'animalcules', and of tiny, complete
animals packed into eggs (preformation). Over time, optics improved. The
reports got better and you could begin to trust the results.

I'm old enough to remember (this is analogy 2) the early transmission
SEM work and how people responded to it. One anatomist told this student
"I believed it until I saw the actual SEMs."

Unkind analogy from my early days as a biochemist: For several decades
surrounding the turn of the 20th century, the research program in
leading biochemical labs was quantitative inorganic analysis. Numerous
papers from that time carried huge tables giving the elemental
compositions of tissues from all sorts of plants and animals. Why?

The reasons were partly sociological. Quantitative analysis generated
numbers, often with lots of decimal places. Biology could now be seen as
just as scientific as physics, or "proper" chemistry. Quantitative
analysis dignified biology and its practitioners.

Second, it was an open-ended program. After you'd analysed liver vs
kidney in sheep, you could do the same in cows, pigs and humans.
Further, you could return to sheep liver vs. kidney when analytical
methods for manganese or other trace elements had improved: better
elemental resolution.

Third, the enormous effort put into the inorganic analysis of Life meant
that vast quantities of data had been published which could be reviewed
and compiled into books: "metapublishing" which gave even greater
dignitas to the leading authorities in the field.

Looking back on this from the 1960s, it was hard for me to understand
how intelligent scientists had got it so wrong. Inorganic composition of
life was a dead end. You could write eminently plausible arguments about
it in 1890, e.g. "There are no vital forces. Life is just chemistry.
Chemistry is the elements. If we learn more about which elements are in
which tissues, and in which proportions, we will move much closer to an
understanding of how Life operates." But you would be wrong, and very
badly wrong at that.

Well, that's what molecular phylogenetics looks like at the moment:
lousy optics, poorly understood statistics and overweening pride. Give
it time, John.

I recently reanalysed a GenBank set of 20 sequences. I used 3 maximum
likelihood methods, 1 Bayesian inference, 2 maximum parsimony and 3
distance-based, and got 14 different trees, all different from the one
published using a 3rd maximum parsimony method. What's going on? Is it
all just nutty?

No. There's a lot of non-phylogenetic noise in those 20 sequences, and
*any* method struggles to get a signal out. Better now to reexamine the
morphology or play with a few more genes, or both.
Dr Robert Mesibov
Honorary Research Associate
Queen Victoria Museum and Art Gallery
and School of Zoology, University of Tasmania
Home contact: PO Box 101, Penguin, Tasmania, Australia 7316
Ph (03) 64371195; 61 3 64371195
Webpage: http://www.qvmag.tas.gov.au/mesibov.html

More information about the Taxacom mailing list