Reply to Mike Zwick

Xuhua Xia xia at cc.umanitoba.ca
Thu Apr 28 11:25:39 EST 1994


In <mezwick-280494060936 at pbmac-11.ucdavis.edu> mezwick at ucdavis.edu (Mike Zwick) writes:

>In article <2pnbdl$omt at canopus.cc.umanitoba.ca>, xia at cc.umanitoba.ca (Xuhua
>Xia) wrote:

>> Such a guess may be thought to be trivial, but it is not. There are
>> 50 or so genes in the phage DNA, including a number of genes that
>> make the head proteins and a number of genes that make the tail
>> proteins. Why should not the genes for the head and tail proteins
>> be expressed first? It is optimality models that allow us to
>> predict, a priori, that head and tail proteins should not be made
>> before DNA replication because otherwise it would be like buying
>> many gift wrappers without buying any gift.

>This simply doesn't logically follow - one can easily imagine a scenaro
>where the head and tail proteins get translated first in order to provide
>the proper context for DNA replication.  It is not optimality models which
>drive the predictions, rather someone just did the experiments and
>determined the answer and made the optimality model explanation afterwards.
> There are many examples of molecular processes in evolution that do not
>appear to be optimally designed (i.e. as a human engineer might design a
>system).

The trick here is that the phage may not be able to carry on DNA
replication and instead may take the lysogenic pathway. In this case
all the head and tail proteins synthesized would simply represent a
burden on the bacterial host. This burden will reduce the fitness of
the phage because its fitness in a lysogenic cycle depends on the fitness 
of its host). If the host is burdened, it will replicate slower, so will
the phage DNA.

>> (I should memtion that Darwinian theory of evolution, with its
>> conclusion that all organisms in nature should evolve towards
>> maximizing fitness under natural selection, is essentially a
>> justification for the use of optimality models in biology. 

>This is absolutely not correct.  While natural selection may cause
>populations to evolve towards some optimum peak on the fitness surface, it
>is not clear that they will ever reach that peak (and thus fulfill the
>prediction of the optimality model).  Thus natural selection is not a
>"justification" for use of optimally models - indeed one can easily imagine
>temporally or spatially varying selection models that never reach the
>optimum phenotype.

This represents a  misconception of optimality models, i.e.
optimality models assume optimality without constraints. In fact, the
appropriate use of optimality models always depends on one's ability
to identify the constraints. What does Mike mean by "optimum phenotype"?
There is no optimum phenotype without specifying constraints. If you
have to define such an optimum phenotype, then it is one that can 
increase the fitness of its underlying genotype at an infinitely large
rate. Neither natural selection nor optimality models requires such
an optimum phenotype to work.

Xuhua Xia
University of Manitoba
xia at ccu.umanitoba.ca

>-- 
>mike zwick
>mezwick at ucdavis.edu
>Department of Ecology and Evolution
>Center for Population Biology
-- 
Xuhua Xia
University of Manitoba
xia at ccu.umanitoba.ca



More information about the Mol-evol mailing list