sibbald at qucis.queensu.ca
Sun Aug 6 10:00:29 EST 1995
| I'd be interested in why you reject the alternative hypothesis that the
| doublet frequencies are controlled mainly by mutational processes (which
| are the same across the genome). For example, a particular doublet
| might mutate more often due to some physical mechanism, and thus its
| steady-state frequency would be lower.
| I would think that if "junk" DNA has a structural role, it would be
| expected to have different doublet frequencies than coding and
| regulatory DNA, not the same ones.
| Mary Kuhner mkkuhner at genetics.washington.edu
First of all, are mutational processes the same across the genome?
If by a "mutation" we mean a mutation that is accepted, i.e. the
organism survives, the answer is no. There is plenty of evidence
for "hotspots" where mutation seems more frequent. There are also
quite a few studies that show in protein coding regions the different
codon positions (1,2,3) mutate differently. It may well be argued
(correctly, in many cases) that raw (not necessarily accepted)
point mutations occur "randomly"
in the genome, and any biases observed are due to selection after
the initial "raw" mutation event. This seems to imply that the
doublet frequencies, even if they arise in the same way, need
not be the same throughout the genome since selection can (and does)
act on the raw mutations filtering out undesirable ones.
It might be argued that the doublet frequencies are just irrelevant
and there is therefore no selection upon them and the raw ones,
all created by the same process are what are observed. Against this
view is the fact that doublet frequencies in different organisms
are quite dissimilar even though i would speculate that the
underlying chemistry and mutational mechanisms (at a low level)
are similar. It therefore seems likely that doublet frequencies
are constrained to be a particular way in a given organism. What
i wanted to know when i studied the different categories of DNA
is if it was the function of the DNA that influenced the doublet
properties. In a word, the evidence for that is weak. So it seemed
that there is something about DNA that requires the doublet
frequencies to be constrained but it is not that the DNA must
code for protein, RNA etc. Since all DNA must be replicated,
remain untangled in the cell and so on, i guessed that the
doublet frequencies are playing some kind of structural role.
Since the frequencies are the same in junk DNA, it seems likely
that the DNA plays some kind of structural role. If junk DNA
was truly junk, it would be unnecassary to ensure that it was
replicated properly, or that it was replicated at all
I hope this addresses the question and apologize for the length
of this reply
More information about the Mol-evol