>>I don't think that this solves the probability problem. Although several
amino acid sequences may function perfectly well for a given purpose, the
number of usable sequences for the given function is not likely to be
sufficiently large so as to deflate the probability argument. <<
I agree that the probability argument is not solved. And at this point, I do
believe that God alone can start life. But I am not about to stake my
theology on that issue because I think that the form of a solution may be in
sight. I have often cited the work of Gerald Joyce and other researchers in
the area of directed evolution. They start with a randomized or partially
randomized sequence of RNA of lengths such that if only one sequence was the
solution the odds against finding that solution are so great that they should
never find it. The fact is that they find solutions to their problems.
Take cytochrome C. It is about 110 amino acid units long. The normal way
that the odds against finding a functional cytochrome C is to say there are
20 amino acids and 110 positions giving 20^110 = 1 x 10^143. Yocky,
(Information Theory and Molecular Biology, (Cambridge: Cambridge university
Press, 1992) p. 254 ) incorporates all the functional sequences of
cytochrome C and drastically reduces the odds against finding a functional
cytochrome. Instead of 10^-143, Yockey finds the odds against finding a
functional cytochrome C at 10^-44. This is a tremendous reduction in
probability.
And interesting sidelight. I had a very short e-mail discussion with
Yockey. I had asked him to arbitrate a disagreement between Brian Harper and
I on the meaning of a certain part of his book. During the exchange, he
asked me what I thought of his probability calculation. I told him that I
thought it was great but that the weakness I saw was that he assumed that no
other sequences other than those in the cytochrome family could perform the
function of cytochrome C. In other words he was equating function with
family and I was not sure that could be stated firmly with the present
evidence. Unfortunately, Yockey chose not to respond to my question as to
how he answered that objection. This means one of three things: It was an
incredibly stupid question which didn't deserve an answer; He had other
things to do; or he didn't have an answer.
But an analogical example of what I am talking about are the following
two sentences.
All nose-pickers get warts on their hands
and
People who stick their digits into their nasal passages develop dermal
growths.
These two sequences of letters are quite different but perform the same
function of warning you what will happen if you pick your nose. They come
from two different families of solutions. To me, the hole in the probability
argument which must be mathematically closed in order to have a firm argument
from probability is this one. "Do no other families of sequences perform the
same function?" Some of the directed evolution work would suggest that they
do.
Steven wrote:
>>However, the issue that no one addresses is how many chances were available
to form highly "improbable" sequences. A commentary based on probability is
meaningless unless one knows how many attempts were made.<<
The number of trials is only meaningful when one knows how many solutions
there are to the problem. If I fill Texas with silver dollars to a depth of
three feet and paint one dollar red and give you one draw blindfolded, from
this pot, you will not pick a red one. But if I paint half of the dollars
red, you have a 50-50 chance. The important piece of info is the phase space
of the solution not necessarily the number of trials.
glenn