Re: Mitochondrial Eve
Sun, 16 May 1999 13:02:13 EDT

In a message dated 5/16/99 1:16:51 AM Mountain Daylight Time, writes:

> Many statistical studies based on mutation rates have *assumed* a constant
> mutation rate. A ratio of eight to one is not even close to constant. This
> suggests (to me at least) that any study which must depend on an assumed
> uniform mutation rate needs to viewed with considerable skepticism.

There are two concepts that may be being confused here. The first is the
rate of mutation versus the rate of gene substitution. Assuming for the
moment that the rate of mutation is constant (I will come back to this in a
moment) the rate of substitution (the number of mutants reaching fixation in
a unit of time) can vary depending on whether the mutation is neutral or
advantageous, and if the latter then it depends strongly on the selective
advantage of the mutation and the population size. As such, it is possible
for different parts of a gene to have very different rates of substitution,
thus making it appear as if the gene has a variable mutation rate when in
fact it only has a variable substitution rate.

The second confusion stems from the fact that while the mutation rate (the
number of mutations that actually occur in a unit of time, whether they
become fixed or not) has been shown to be more or less constant between
different regions within a gene, it has also been shown to vary between
genes, sometimes enormously (though for each gene its own rate is still more
or less constant). In other words, mutation rates are like the rates of
radioactive decay. While the rate of decay is different for each isotope,
the rate for any one isotope is constant; similarly the rate of mutation is
different for each gene, but the rate for any one gene is constant. It
sounds from the Hagelberg et al quote that the authors are describing the
mutation rate of a different gene (or portion of a different gene). In other
words, one hypervariable region of one mitochondrial gene has one rate of
mutation while another hypervariable region of another mitochondrial gene has
a rate eight times faster. Yet both rates are more or less constant. See
_Fundamentals of Molecular Evolution_ by Wen-Hsiung Li and Dan Graur (Sinauer
Associates, Massachusetts, 1991) for more details.

There is also possibly a third confusion, that of mutation rate versus the
rate of molecular evolution. The former is strongly experimentally
demonstrable and so generally accepted; the latter has less experimental
support and so is controversial. So we need to determine exactly what the
authors are referring to before we can make any conclusions about the
validity of statistical studies based on mutation rates. Are they talking
about the rate of mutation, the rate of substitution or the rate of molecular

Kevin L. O'Brien