Re: Poisson Distribution

Seanpit wrote:
On Mar 20, 9:13 am, John Harshman <jharshman.diespam...@xxxxxxxxxxx>
wrote:
The reason why HW has its own branch is because it has a sequential
mutational difference compared to all the other branches. This
difference is minimal compared to HB and most significant compared to
HE (i.e., more mutational differences).
I have no idea what "sequential mutational difference" means. Could you
present a tree that's not garbled? Then I might be able to figure out
what you mean.

1. AAAAAAA
2. BAAAAAA
3. CAAAAAA
4. BBAAAAA

Sequence 1 is the same "distance" from sequences 2 and 3 (i.e., 1
difference), but is farther away from sequence 4 in sequence space
(i.e., 2 differences).

So, by "sequential mutational difference" you merely mean "genetic distance". OK. Still don't know what your example tree is supposed to look like.

I'm not sure what you are talking about here. Perhaps a diagram will
help to illustrate what I'm talking about a bit more:
http://www.detectingdesign.com/images/EarlyMan/Neandertal-Human%20Phy...
That diagram, simplified as it is, doesn't even say what you think. Now
in fact the greatest difference among humans is between two points you
don't label, in the "northwest" and "southeast" of the cloud.

The cloud is essentially circular. In fact, statistically, it should
really be drawn in a near-perfect circle.

Statistically, there is no such thing as this "cloud". There is a tree. Even the tips of the tree will not form the sort of cloud you show here.

This shows
that, even in two dimensions, your claim that the greatest difference
among humans need not contain the sample with the least difference
between human and neanderthal.

The claim is that the closest modern human sequence to a Neandertal
sequence will very likely be farther away from at least some other
modern human sequences. That's the claim. That claim is very much
supported by the diagram and the statistics of likely tree twig
distributions in n-dimensional sequence space.

Sorry, my sentence appears to have been garbled. Your claim is that the human closest to neanderthal will also be the sequence having the greatest pairwise distance between two humans. Right? That claim is false. The diagram doesn't support it, even as drawn, and you don't understand the statistics. What your claim would seem to require is convergence between humans and neanderthals. If indeed human sequences are intended to be diffusing from the center of the circle toward the edges, what you claim is that this diffusion will bring some humans closer to neanderthals than the common human ancestor. This is so statistically unlikely as to be unworthy of serious consideration. And your two-dimensional representation is at fault here; convergence would be fairly likely with a 2-dimensional (two-base) sequence space. But the probability decreases rapidly as the number of dimensions goes up.

And the diagram is bogus in two other
ways. First, it's not a tree, which mitochondrial differences most
certainly are.

The diagram represents a 2D projection of n-dimensional sequence space
- a space that is occupied by the twiggy tips of a tree that branched
over time.

In that case, it's a false representation of the space.

Second, it's two-dimensional, when mitochondrial
distances would be n-dimensional, where n is the length of the sequences
being compared. I find it amusing that a bogus diagram still fails to
convey your point.

Obviously, the diagram is a 2D projection of n-dimensional space (very
similar to the 3D projection of n-dimensional space illustrated by
Choi and Kim in their paper concerning the various relationships of
single protein systems).

Sadly, the relationships among sequences in your 2D projection has nothing to do with the actual relationships among DNA sequences. Convergence, which you assume, is just not going to happen. Sequences get farther apart during evolution, not closer together. Convergence at individual sites is likely, but this doesn't add up to convergence of sequences.

.

Relevant Pages

  • Re: Humans, chimps, wheat and frogs
    ... conserved) sequences are a tiny minority of the genome, ... These stretches of DNA ... to hybridize between chimps and humans is the result of missing DNA ... As the species specific miRNAs described here are ...
    (talk.origins)
  • Re: Humans, chimps, wheat and frogs
    ... conserved) sequences are a tiny minority of the genome, ... These stretches of DNA ... Humans and chimps differ by 40 million mutations. ... As the species specific miRNAs described here are ...
    (talk.origins)
  • Re: Of Mice and Straw Men
    ... > jump to infecting humans. ... > sequences declines exponentially with increasing minimum size and/or ... Higher levels are connected to lower levels by these ... > So, you see, with higher and higher levels of functional complexity, ...
    (talk.origins)
  • Re: differences between chimps and humans
    ... compared to humans. ... species differences should show the fastest evolution ... You say "sequences having the greatest effect on species difference ... greatly between chimps and humans but hardly at all between chimps and ...
    (talk.origins)
  • Re: Gutierrez et al., make same mistake as Sean Pitman
    ... modern humans were more similar to Neandertals than to other modern ... selected sequences and the Neandertals rendered 1.6% of human-human ... This assumes that distances are independent of each other. ...
    (talk.origins)
Loading