Lucy in Newsweek

[HJJACOBSON at AOL.COM]

In a message dated 4/1/2004 9:04:47 AM Pacific Standard Time,
jcclark at CSUPOMONA.EDU writes:

> That's one way of doing it. Maximum parsimony, with outgroup comparison, is
> another. A base position is just another four-state character in that sort
> of analysis. And then there's maximum likelihood, that many also claim as a
> cladistic technique.
>

I'm not certain that a base position is "just another four-state character."
Are you telling me that a particular base in a particular location can be used
to identify a clade? If that were the case why not just compare sites instead
of calculating coefficients? I can see considering the sequence as a single
character and variations in the base positions as states of that charter.
Hence, certain sequences could be considered apomorphic and used to identify
clades.

As for maximum likelihood all I can say is a lot of people claim that a lot
of methods are cladistic. That was the point of my response.

As for substitution rates being clock-like, perhaps I used the wrong term
here because I thought that the reason for using conserved sequences was that the
only differences were due to background mutation which were clock-like.
Should I have said that the mutation rate was clock-like?

You are correct NT never required clock-like mutation rates. In fact it
didn't require any assumptions about evolution, which why I always found it
intellectually honest. And NT never claimed its dendrograms were phylogenies because
it was impossible to tell if the mutation rate in the study group was
clock-like, unless, of course, you knew the phylogeny of the group before you started.
My point was that clock-like mutation rates are necessary if overall
similarity is used to estimate phylogney.

Herb Jacobson