How does recombination increase information in the gene pool?
By itself, one simple mechanism - asymmetric recombination resulting in gene duplication.
This, then allows for more interesting point mutations - one copy continues to produce a product that fulfills the original gene function, the other can accumulate point mutations.
I didn't say point mutations are not involved in creating diversity, just that I don't think it's as important as other mechanisms.
Or genes can be aquired from another genome via heterologous recombination. Now you have a gene product in a different environment. That may not add information, but change the interpretation of the same information (i.e. pleiotropy).
Don't get so wrapped up in genome and forget that selection operates on the proteome; there's a lot of information that's not directly readable from the DNA sequence.
Recombination does allow for diversity in a population but recombination (without errors) and natural selection can only cause loss of information in the gene pool you can never create a new gene by this mechanism.
Why the qualification? In the real world, does recombination occur without errors?
Of course you can create a new gene by recombination. The eukaryotic genome is inherently modular (you know, introns and exons); recombination can move pieces into and out of genes, creating new genes. I should try to find a few specific examples, but it's been a good 10 years since I was taking molecular biology coursework, so I don't have any in mind at the moment. Sorry, but I'm getting ready to be traveling tomorrow.
Then there are the transposable elements. See shapiro.bsd.uchicago.edu/Shapiro_1999_Genetica.pdf for a review (that name brings me back - I think I exchanged few emails with Dr. Shapiro a few years back; I seem to remember he was misquoted by some anti-evolutionists).
kleinman said:
So let’s do a little arithmetic. Assuming the length of the human genome is 3 billion bases. 2% of 3 billion is 60,000,000 base differences. If the humans and chimps diverged 4,000,000 years ago and you assume an average generation time of 10 years give 400,000 generations to accomplish this 60,000,000 base differences. That works out to an average of 150 nonfatal mutations per generation.
You're assuming point mutations only.
Consider that a good 10% of the human genome is composed of repeated Alu elements - roughly 300bp of retrotransposable DNA that serves no known function, but can jump and replicate relatively freely. Consider w w w.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=10381326&dopt=Citation
A single Alu, differing from the original by a single point mutation, amplifying itself 1000 fold over a single generation would not shock me in the least (note that the above reference quotes an Alu copy number of >500000, and one new insertion for every 200 new births).
Also consider the role of retroposons. See biology.plosjournals.org/perlserv?request=get-document&doi=10.1371/journal.pbio.0030357
A little arithmetic shouldn't replace a lot of biology.