On Thu, 23 May 1996, Ludvig Mortberg wrote:
> The reason I ask is that I have, unsucessfully, been trying to get a
> hold of what the clock is all about. Some people tell me that constant
> change has been abandoned years ago and others that it is still the
> foundation of molecular systematics.
Both those claims are correct. (Sorry, couldn't resist that set-up ;-))
But seriously Ludvig, here are my two cents: To a first approximation most
genes seem to evolve in a clock-like manner, inasmuch as the strength of
purifying selection acting on each gene is constant through time. This
claim was codified in the inter-species half of the neutral theory, due to
King and Jukes. And this assumption underlies molecular systematics.
When don't genes show clock-like evolution? Whenever natural selection
rears its ugly head. Thus balancing selection gives rise to ancestral
polymorphism and wrong trees (see Ayala 1995, Science 270:1930 for a
cannonical example: MHC in humans and chimps) and directional selection
would temporarily speed the clock up (if it ever happened).
Some other potential (nonselective) complications with the clock include:
change in generation time, if mutation rate is constant per round of
gametogenesis rather than per trip 'round the sun (Li et al 1996, Mol
Phylog and Evol 5:182); comparison of sex-linked genes, if mutation rate
is higher in spermatogenesis than oogenesis (Chang & Li 1995, JME 40:70);
change in population effective size, if purifying selection is Ns but only
s is constant (due to Tomoko Ohta); and lineage-specific change in
mutation rate (Martin, Naylor and Palumbi, Nature 357:153).
Doubtless there are other problems. IMHO, the clock is a naive model
which nevertheless accounts for a lot of data, but to which there are
loads of very interesting exceptions.
Daniel M. Weinreich email: dmw at mcz.harvard.edu
Harvard University usmail: 26 Oxford Street
Museum of Comparative Zoology Cambridge, MA 02138
voice: (617) 495-1954 fax: (617) 495-5846