Re: Moving the Islands within Sequence Space

On Jan 25, 2:50 pm, seanpitnos...@xxxxxxxxxxxxxxxxxxxxxxxxxxx wrote:
On Jan 25, 11:29 am, Garamond Lethe <cartographi...@xxxxxxxxx> wrote:



On Sun, 25 Jan 2009 11:08:09 -0800, Seanpit wrote:
On Jan 8, 11:54 pm, "R. Baldwin" <res0k...@xxxxxxxxxxxxxxxxxxxx> wrote:
Seanpit <sean...@xxxxxxxxx> wrote
innews:6cee2bbe-63a6-4a27-

ada0-605ea814f...@xxxxxxxxxxxxxxxxxxxxxxxxxxxx:

On Jan 8, 7:33 pm, "R. Baldwin" <res0k...@xxxxxxxxxxxxxxxxxxxx>
wrote:
Seanpit <sean...@xxxxxxxxx> wrote in news:2fdb2619-c052-4fda-ac3b-
97f1505d9...@xxxxxxxxxxxxxxxxxxxxxxxxxxxx:

On Jan 7, 9:00 pm, "R. Baldwin" <res0k...@xxxxxxxxxxxxxxxxxxxx>
wrote
:

You have no clue what you're talking about.  The Poisson
distribution isn't a "process".

Please look up "Poisson process" on wikipedia and stop
pretending that you know what you are talking about.

I'm not using a "Poisson process" in my arguments.  I'm using a
Poisson distribution.  Look that up.

Sean, I suggest you follow Mr. Stockwell's suggestion and read
about
Poisson processes. Mr. Stockwell did not tell you that a Poisson
distribution is a Poisson process. If you read about Poisson
processes,
then you would understand what he meant, and learn something to
boot.

From Wikipedia:  "A Poisson process, named after the French
mathematician Siméon-Denis Poisson (1781 – 1840), is the
stochastic process in which events occur continuously and
independently of one another (the word event used here is not an
instance of the concept of event frequently used in probability
theory)."

What I'm describing is only a Poisson process in the sense that
random mutations do in fact occur continuously and independently
of one another in different individual organisms and even
different individual sequences.

Well, at least this is progress from "I'm not using a "Poisson
process" in my arguments. I'm using a Poisson distribution." Do you
understand yet why Mr. Stockwell corrected you, and that he was
absolutely correct in discussing the Poisson process in your model?

It wasn't apparent to me that he was actually discussing the Poisson
process of random mutations.  That's obvious.  What is completely
missed is the fact that the odds of hitting upon a target are in fact
dependent upon the ratio of targets vs. non-targets.

Yes, we know, if you are searching a sparsely populated large space
with a random search.

Other than that, my description of the odds of finding rare
targets with unknown locations in sequence space by a process of
random mutations seems to be completely missed by both you and
Stockwell.

No, we haven't missed your description. We've tried to explain why
it does not agree with reality (so have others), but it appears you
haven't been willing to listen.

You have tried, unsuccessfully, to challenge my math.

I have more than once successfully corrected your math, and you've
acknowledged it at least twice, as you might recall. I made a mistake
in this thread and self-corrected it on the same day..

Tell me, what
else do you have?  You won't present your own alternative model
because, as you say, you aren't a biologist.  You really have no idea
how RM/NS is capable of doing the job.

Actually, I do have an idea, by reading articles by biologists.

Yet, somehow, you just know
that my model is flawed.  How, pray tell?  Please, do present
something relevant this time.

Because your model arm-waves away fundamental research about how
proteins actually evolve in only two sentences, yet provides no
supporting evidence.

Your papers only talk about protein evolution below 100aa.  My model is
talking about evolution where the minimum requirement is greater than
1000aa.  Have anything relevant to that problem?

Several.  You declined to read them, remember?

You declined to produce a relevant quote from any of them that
described either 1) evolution in observable action or 2) any
statistical analysis concerning the potential for the mechanism of RM/
NS to do the stated job beyond 1000 fsaars.

There is no such paper in all of literature.  All of the papers that
claim that RM/NS works beyond 1000 fsaar are based on nothing but bald
assertions and baseless assumptions.  

Only if you define bald assertions as actual experiment and baseless
assumptions as demonstrating the "possiblity" by pointing out possible
source sequences that *actually* exist in organisms.

There is no statistical support
for these declarations.  

Nope. No bogus numerology based on obviously false assumptions. Just
*real* evidence.

There is nothing but hot air.  It is simply
assumed the RM/NS did the job based on the notion that any degree of
homology, regardless of the actual Hamming distance, could have been
crossed in a few hundred million years by RM/NS.  

Since actual Hamming distances for proteins with the *same* function
is mostly composed of selectively neutral changes, you don't even need
to invoke NS in most cases. Changes due to NS for new (modified,
emergent, additional) functionality would be much more rapidly fixed
into populations.

That assumption is
just that - a bald assumption.  

It is not an assumption that two proteins that have the same
function. In many cases, they can be switched out in actual
organisms, and they certainly are essentially equivalent in the test
tube. And it is not an assumption that the pattern of sequence
changes (amounting to differences in Hamming distances) is a
reflection of the time since divergence from a common ancestor. In
fact, it has strong statistical support. As does sequence similarity
in gene families for genes that perform *different* functions. Those
sequences are NOT the result of independent completely random searches
from some random site in total sequence space to another random site
in total sequence space.

That's not science.  Sorry.

Bogus numerology is not science. Neither are assertions that the only
possible alternative to a process determined to not be completely
random (and, remember that I agree with your rejection of a totally
random search in total sequence space that your math is testing) that
whatever you want was done by some untestable, invisible something
that did something somehow at some place and some time to produce
whatever you see.

Sean Pitmanwww.DetectingDesign.com

.

Relevant Pages

  • Re: Moving the Islands within Sequence Space
    ... Poisson distribution. ... Mr. Stockwell did not tell you that a Poisson ... I'm using a Poisson distribution." ... how RM/NS is capable of doing the job. ...
    (talk.origins)
  • Re: Geraldine bulletin
    ... Costing the net hundreds if not thousands of dollars, Befunge Sudoku ... > Well, when you have a sequence of trials, you have to do the Poisson ... Really, the binomial thing, but it approximates so closely to Poisson ...
    (uk.rec.sheds)
  • Re: Poisson generator
    ... I need to generate a sequence of poisson distributed numbers between ... I have a poisson number generator. ... side by zero but are not bounded on the high side. ... But the OP already has a Poisson number generator. ...
    (sci.stat.math)
  • Re: Poisson generator
    ... I need to generate a sequence of poisson distributed numbers between ... I have a poisson number generator. ... I can post the code for a Poisson random number generator here. ... side by zero but are not bounded on the high side. ...
    (sci.stat.math)
  • Re: Distribution of a vowel on the page
    ... times a vowel occurs on a line follows Poisson distribution or not. ... vowels and consonants in words violate that assumption. ... It occurs to me that if there is a positive correlation of having ...
    (sci.stat.math)
Loading