Code center > Suggestions
Recombination mechanism
Numsgil:
I made a modification to Windiff and put it on the FTP a while ago. It ignores capitalizational differences.
But that's sort of off-topic, isn't it. :rolleyes:
shvarz:
--- Quote ---Only zipped up regions can have a crossing over event occur.
--- End quote ---
This is a possibility... We can align very short stretches of DNA to zip them up. What we can do is this: At some point during copying the DNA, program decides to cross-over. It then takes N next commands (5-10 should be enough; make it user-definable?) and searches for exact same pattern in the DNA of the other parent - if it finds a match, it switches over. If it does not - sticks with DNA from the same parent. If it finds several matches, then it picks one of them randomly.
This would solve the problem of sexual reproduction, it would also allow recombination that is command-oriented, not gene-oriented. The downside is that it still would not be possible to track evolution of whole genomes and build phylogenetic trees. But it would be a good start anyway. It actually would also allow turning off cross-species breeding - just set the "homology" level very high and it will never find exact match :)
Yeah, I kind of like that!
Numsgil:
--- Quote ---If it finds several matches, then it picks one of them randomly.
--- End quote ---
I was thinking of that too, but shouldn't distance from current location matter too?
That is, if there's a closer match you're more likely to use it than a match on the opposite side of the centromere.
shvarz:
More likely? Sure, why not...
Can you explain the whole centromere again or post a link? I am not 100% sure what you want to do there.
Numsgil:
Here's the most recent post.
Any questions that doesn't answer (I don't remember all of what I said) go ahead and ask in that thread.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version