Re: The shape of inbreeding/outbreeding depression


"Robert Maas, see http://tinyurl.com/uh3t"; <rem642b@xxxxxxxxx> wrote in message news:dc77k2$deg$1@xxxxxxxxxxxxxxxxxxxxxx

> > From: Guy Hoelzer <hoelzer@xxxxxxx>
> > Assuming these sequences are properly aligned to ensure the homology
> > between opposed letters, the sequence similarity between these two
> > sequences is 0% because none of the opposed letters are the same.
>
> I don't believe anyone of reasonable intelligence would make such a
> stupid assumption, so you seem to be creating a strawman to knock down,
> so your hypothetical calculation is worthless.
>
> > I think my calculation is clear from my statement, although it only
> > required scanning the sequences and counting similarities.
>
> It's not clear at all what algorithm you use to count what matches as
> opposed to what doesn't match, given the cost of rearranging one or the
> other genome sequence to make it match the other. Is that cost zero,
> whereby any shuffling whatsoever of some mix of the four bases would
> count as identical to the original, or is that cost infinity, whereby
> if there's a single rearrangement then the similarity is considered to
> be zero as you strawman-calculated above? Since you refuse to put a
> number on the cost of rearrnaging to achieve alighment, there's no
> point discussing such matters with you. Your method is non-defined.

Robert,
It would be a courtesy if you did not snip so much context. It took
me some effort (and a change in news viewers) to determine that the
context was as follows:
---------
>> GH:-
>> I would argue that percent sequence similarity utterly unambiguous in
>> terms of measurement
> RM:-
> Only inside your mind, where you apparently have a single definition
> and can't believe anybody would use a different definition. Consider
> the following two sequences of letters:
>
> qwewrtyuiopasdfghjklzxcvbnm
> jklzxcvbnmqwewrtyuiopasdfgh
>
> If you calculate the percent similarity between them, per whatever
> definition you have in your mind, what number do you get as the answer.

GH:-
Assuming these sequences are properly aligned to ensure the homology between
opposed letters, the sequence similarity between these two sequences is 0%
because none of the opposed letters are the same.
---------
If you don't snip the context, it is possible for an observer to
form his own opinion. My opinion, for what it is worth, is that Dr.
Hoelzer is not being stupid - merely tolerably sly and evasive.
Alignment is a bit of a problem for Hoelzer's scheme, but perhaps not
a fatal one. Hoelzer's "evasiveness" shifts the burden of proof (that
it IS a problem) to you - which is where it properly belongs to begin
with.


.

Relevant Pages

  • Re: "Human" defined by Junk (non-Coding) DNA
    ... segments harbour 1,243 non-coding sequences conserved between humans ... So 70% conservation over a stretch of 100 bp sure seems ... And is "pure junk" defined as such because 45% shows ... no sequence similarity, or is there a threshold of sequence similarity ...
    (talk.origins)
  • Re: EOF code?
    ... of the cost of the Christian's method. ... averaged over all bit sequences up to some number of bytes that those ... this time it cost one byte with a probability of ~1/2. ... I found that for his transform, all bit sequences whose lengths were ...
    (comp.compression)