Re: DNA sequencing

Chris Frazier (cfrazie@unm.edu)
Sat, 7 Oct 1995 11:37:53 -0600

O. K., here's a little addition for round 2.

>Have there been cases where the DNA studies have given data that
>is flat out wrong?

Kind of. The hard partof molec. phylogenetics is taking your information
from DNA sequences and converting it sensibly to a pattern of relationship
of the organisms you were looking at. This is not easy and the methods are
not entirely worked out yet. In fact, that talk I refered to on the
relations of Triphophyllum (Dioncophyllaceae) and Ancistrocladus
(Ancistrocladaceae) used a demo copy of a brand new and supposedly much
more powerful technique for analyzing the confidence associated with a
given tree of relationships created from molecular data. I might add that
the tree or relations among the CPs that Fernando was looking at is only 1
of 396 equally good, but different ways the authors found for drawing that
tree. Most of the other trees probably varied in only small respects from
the one they showed, but the authors did just pick one to present in the
paper.
As to flat out wrong, I presume that is what Jan thinks about the
molec. results showing the "close" relationship between the Caryophyllids
(Carnations, Purslanes, Cacti, Chenopods, etc.) and the group containing
Nepenthes,Drosera and Dionaea. I personally think this result fits O. K.
with the morphological data (look at the lack of petals, separate sexes,
modified hairs and the preference for "bad" habitats shown by the Chenopods
for example), but Jan might know more about this than me.

>... even if it's not free from selection, wouldn't selection be
>acting equally on a specific gene common to many organisms or at least
>acting equally on each intron and exon? Wouldn't introns be the perfect
>neutral markers since they don't code for proteins?

Well that's the dogma and its probably more or less correct. But it is
probably wrong to think that introns are neutral junk because they don't
code for proteins. We don't currently think they have a coded function,
but they probably do, we just don't know what it is (maybe they code for
enzymatic RNA, who knows). Certain proteins tend to be used a lot in
molecular phylogenetics because they are thought to be more or less
neutral, but in one case, for example, I have seen a paper that suggested
that different forms of a commonly used protein are related to differences
in water uptake ability. So with this protein you may be looking at
ecology not taxonomy. Determining how strongly selection is acting on a
character is difficulty for observable traits and it is just as hard for
molecular traits. And no, I don't think there is any reason to assume that
just because a trait is at the protein or DNA level that selection is
acting on all species the same. To some degree we are all just as
different inside as out.

SIDE NOTE:
I would like to repeat my request for any information concerning attempted
Nepenthes crosses that didn't work. I don't know whether to interperet
lack of responses so far as indication that Nepenthes species are
completely interfertile as far as anyone out there knows or if it just
reflects the shyness of CP growers to admit their crosses failed : ) (Shy
folk can send their info to me directly, I won't look at the "From:" line).

Chris

-----------------------------------------------------------
Chris Frazier
Dept. of Biology, Castetter Hall, UNM
Albuquerque, NM, USA 87131

Wk: (505) 277-0683
Home: (505) 255-2176
E-mail: cfrazie@unm.edu