robison at lipid.harvard.edu
Thu Nov 16 08:15:27 EST 1995
Ben S. Hadad (hadad at Angelique.org) wrote:
: I have what is, I hope, not too obvious a question...I'm a programmer
: working with genetic algorithms, and am wondering if, in real life,
: crossover breaks a chromosome in a completely random location, or
: whether there are some favored spots...Is there any sparing of
: 'genes', codons, whatever, from splitting, or just willi-nilli?
Yes, there are favored spots, but they generally represent
some sort of property of the overall DNA sequence and not
some sort of reading of the DNA's meaning at that point
(though it is possible that evolution might select for/against
recombination at particular spots). Perhaps the best studied
is the chi site in E.coli, a polar 9 base pair site which stimulates
recombination. Interestingly, chi sites are found in a highly
biased polarity distribution relative to the direction of DNA replication.
There really isn't much reason _a priori_ to expect recombination from
occuring amidst a codon. Remember, the bulk of recombination involves
exact pairing between two pieces of DNA coding for about the same
thing, and so there are unlikely to be harmful effects from
the recombination event (i.e. recombination is unlikely to generate
a frameshift from two unframeshifted genes).
Good genetics or molecular genetics textbooks discuss recombination
in some detail. You might try "An Introduction to Genetic Analysis",
"Genes V", or "Molecular Biology of the Gene".
Department of Cellular and Developmental Biology
Department of Genetics / HHMI
robison at mito.harvard.edu
More information about the Mol-evol