How do you evolve an organism?

Here anything Cell Lab related that doesn't fit into the other topics can be discussed.
Post Reply
User avatar
RedShift
Posts: 43
Joined: Mon Dec 17, 2018 10:04 am

How do you evolve an organism?

Post by RedShift » Mon Sep 02, 2019 11:22 am

I'm just curious, how do you evolve an organism. Like what amount of radiation do you use per substrate age/hour. I used to set the radiation to max and after an hour set it to 0 and leave it in 0.50 — repeat. :?:
Uhmmmm...
User avatar
countingtls
Posts: 25
Joined: Mon Jun 10, 2019 10:36 am

Re: How do you evolve an organism?

Post by countingtls » Sat Sep 14, 2019 12:35 pm

RedShift wrote:
Mon Sep 02, 2019 11:22 am
I'm just curious, how do you evolve an organism. Like what amount of radiation do you use per substrate age/hour. I used to set the radiation to max and after an hour set it to 0 and leave it in 0.50 — repeat. :?:
(I think you missed a 0 after the decimal point, the radiation level range is between 0 and 0.090)

After a lot of using pure evolution to finish challenges, I feel it depends on what's the goal. If you are trying to evolve new structures/modes (not single point mutation), the radiation level during the mutation period can be pretty high. It depends on how long you want to wait generally beyond 0.005, at first everything is random the higher the better, but in later stages, lower values need to be used without causing extinction (unless you want very simple and robust organisms). But they have to be solidified/trimmed when certain new structures/species evolved, hence during these solidify period set the radiation level to 0 if you only want one species as the final result.

If the goal is to optimize existing organisms (single point mutation), then it depends on how sensitive the organism is to slight changes in tuning parameters. Say if you have a very complicated smart swimmer/hunters depends on very delicate signal values to function, and complicated growing steps lead to the birthing of the next-generation organism, then probably a very very low mutation level (like 0.001 to 0.010) is preferable, and the periodic solidify is usually not necessary but only once at the end to give you one stable solution. Or if the organism is robust/simple enough, a much higher rate like 0.020 to 0.030 is still doable.

For evolving multiple species as an ecosystem with a constant mutation rate throughout, then I feel a level beyond 0.005 even to 0.020 is OK if you are willing to run simulations for actual hours and days. Higher-level like beyond 0.030 will generally break the ecosystem and lead to monotone species
Post Reply