This is a fascinating ongoing discussion from the internet, on DNA and evolution.
Kendell Hyde quoted: "The rate of base substitution in mtDNA is much higher than nuclear DNA. An estimate of the initial rate of sequence divergence is 20x10-9 per site per year per evolutionary line (this is 2% sequence divergence per million years between pairs of lineages; 10 times faster than the highest rates in nuclear DNA)."
Restating this number for nuclear DNA, the base rate substitution is about one per line per year (actually 1.5, considering 3x109 bases). Rich, I am trying to be very clear about this number, which I have seen several places.
DHB quoted several days ago that human and gorilla beta hemoglobin are 99.3% the same. The beta hemoglobin has 147 amino acids, or a little over 400 bases. .7% difference would be 2 or 3 bases. Given the number of humans, each year every DNA base could be changed by the process above more than once, so any single change would be easy, somewhere within the human population. It is the second change that requires some thought.
Most mutations in working DNA, the exons and control sequences, are thought to be fatal, and some of those sequences are highly conserved, like being the same in yeast and Homo. However, just looking at the base rate substitution as above, the probability of another change in the 400 base beta hemoglobin chain is 400/3x109 per year. If that change has to be a specific base, and a specific type of substitution (of A, G, C, T) the chance becomes 1/4 per 3x109 per year, or about 10 billion years to get the correct 2nd substitution. If there are more than one person who got one correct mutation, then the probability goes to maybe several in 10 billion years. However, it has only been several million years since Homo and gorilla diverged. And if there are 3 bases to be changed by random mutation, that multiplies the probability by about 1/10 billion, or 1 chance every 1018 years.
The more we find out about exactly how the genome works, the more difficult it is to explain how natural random processes could be the source of evolution.
I was reading in the current Sci Am, where an article explains the way cells produce proteins from DNA. It says that there are about 150 proteins involved in the splitting and copying of the DNA. I would guess that a cell without those proteins available would not be able to reproduce any proteins. I would think that the sperm and egg contain a set of those copying proteins. I understand that some of the very simple single celled organisms use a simpler copy method, but it would seem that even they have to have the necessary support system to copy any DNA. Makes me wonder just how complex the first cell would have to be in order to make the first copy.