On Jun 21, 7:59 pm, David <david.neff1...
OK, I am sure I have errors in this... but here is my shot at it...
From the article:
Mutations of any kind are believed to occur once in every 100,000 gene
replications (though some estimate they occur far less frequently).
Davis, 68; Wysong, 272. Assuming that the first single-celled organism
had 10,000 genes, the same number as E. coli (Wysong, 113), one
mutation would exist for every ten cells. Since only one mutation per
1,000 is non-harmful (Davis, 66), there would be only one non-harmful
mutation in a population of 10,000 such cells. The odds that this one
non-harmful mutation would affect a particular gene, however, is 1 in
10,000 (since there are 10,000 genes). Therefore, one would need a
population of 100,000,000 cells before one of them would be expected
to possess a non-harmful mutation of a specific gene.
The odds of a single cell possessing non-harmful mutations of five
specific (functionally related) genes is the product of their separate
probabilities. Morris, 63. In other words, the probability is 1 in 108
X 108 X 108 X 108 X 108, or 1 in 1040. If one hundred trillion (1014)
bacteria were produced every second for five billion years (1017
seconds), the resulting population (1031) would be only
1/1,000,000,000 of what was needed!
Why is he calculating the odds of one non-harmful mutation in a
*particular* gene? A non-harmful mutation in any gene is non-harmful
regardless of which gene it occurs in, and could thus be transferred
to subsequent generations, which is all that should be required for
the mutation to survive. Therefore, one would need only a population
of 10,000 cells rather than 100,000,000 cells for one of them to be
expected to possess a non-harmful mutation of any gene.
Why is he calculating the odds of a single cell possessing non-harmful
mutations of five *specific* (functionally related) genes? Firstly,
what does functionally related mean? Secondly, the base pairs should
be considered rather than the genes, since genes can arise out of non-
code junk. But in any case, a set of non-harmful mutations in *any*
set of "functionally" related genes is all that should be required for
a viable "functional" change. The result should then be the product of
their separate probabilities (lower than suggested in the article as
described above) multiplied by the combinatoric number of sets of
functionally related genes that might exist (of any number, not just
of sets of 5). How many permutations of genes are in the Genome size
of E. coli (about 4.7 million base pairs and about 4,400 genes) could
be "functionally" related? Nobody knows that. Lets guess it might be
as small as 10,000 (please suggest a better number if you have one).
Assuming 5 was correct for the average number of "functionally"
coordinated genes sets (horseshit, I know), we have 1 in 10^4 * 10^4 *
10^4 * 10^4 * 10^4 * 10^-4 or 1 in 10^16 (not 1 in 10^40).
Why is he taking a population of only one hundred trillion E. coli
(10^14)? That is about how many there are in a single human gut. And
E. coli is not the only bacteria. World wide E. coli is in the order
of perhaps10^30. The bacterial count is ??? And why the constant
growth rate when it should be exponential growth up to the point of
full worldwide colonization, quickly bringing any population to
current world levels?
Actually bacteria duplicate about once every minute rather than once
per second, so I will cut him some slack on that. But it would still
give 10^30 / 10^16 = 10^14 functional mutation per minute worldwide.
I may have errors, and I welcome corrections. But lets face it. The
original article is to freaking stupid for words.