Code center > Suggestions
Recombination mechanism
shvarz:
--- Quote ---f you can find a link on how real organisms line up their chromosomes (I looked but didn't find anything), that would rock!
--- End quote ---
As far as I can tell, this is still unknown! The general feeling is that DNA sequences are somehow aligned alonside each other in a very accurate manner (down to a base). We know only one way to accurately align DNA strands - unwind the double helices and allow base-pairing of homologous regions (do you understand me here? sounds a bit too technical). There is some evidence that this happens, but no definite proof. Google for synaptonemal complex and recombination to find some pages that talk about this.
shvarz:
--- Quote ---Which is a blatant violation of one of the principles of DB, that is, that species name is just a comfortable label and has no meaning inside the simulation.
--- End quote ---
Sometimes you don't want it to be just a label. Say you want to compare fitness of two closely related organisms that reproduce sexually - then you don't want them to inter-breed.
--- Quote ---there are standard and well known routines to calculate the common parts and similarities between to texts. And they are relatively fast.
--- End quote ---
Well, in my lab there are people who study evolution of viruses and they often need to build phylogenetic trees for them. They are using the best software available and thier alignments usually take overnight and sometimes go for several days. Which makes me think that any program that does this in a fraction of a second is a sham. Well, maybe it can do some half-assed job, but I am sure it will stumble over some of the things that evolution does.
BTW, I don't know about text editors that programmers use, but Word gets totally fucked up when you try to compare documents with several parts moved around. I tried it, it does not work :)
Basically, if you can do it - great! I won't complain. But I don't see why UID system can't be implemented - it is very powerful and yet it is very easy to add...
Numsgil:
Okay I understand the problem a bit better now.
A ID system would work, but you're right, it makes DNA messy. Instead we could define a particular series of commands, something that's likely to be "Junk", and have the program "zip" up homolgous chromosomes starting at these points. When an inconsistancy occurs, you stop.
Or instead of Junk DNA, we could use a special command. Either works (although a command just for this seems overkill to me).
If you have enough such commands, every inconsistancy is effectively isolated. If not, you suffer from problematic unequal crossovers. Benefit is that the DNA still becomes readable without alot of fluff.
We could even have the DNA 'correct' itself if it finds an inconsistancy. We could also change how large the inconsitancy needs to be to stop the 'zipping' up.
Only zipped up regions can have a crossing over event occur.
That's consistant with everything I understand about crossing over. Tell me how well that works shvarz.
Numsgil:
--- Quote ---Sometimes you don't want it to be just a label. Say you want to compare fitness of two closely related organisms that reproduce sexually - then you don't want them to inter-breed.
--- End quote ---
The best way to solve this problem is to require a kind of 'handshake' signal that both bots want to mate.
I was thinking of implementing sexual reproduction through ties. So both bots would have to set a variable to true to allow the mating to occur. No rape in DB!
--- Quote ---BTW, I don't know about text editors that programmers use, but Word gets totally fucked up when you try to compare documents with several parts moved around. I tried it, it does not work :)
--- End quote ---
He's right, I haven't found a good document comparing utility that can recognize that I moved text segment A from the top of the document to the bottom. Windiff just tells me they're different in those places, which isn't particularly useful beyond a point.
Botsareus:
yea I was looking for a better program then windiff too , no luck yet. I need binary compare though.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version