Tips, tricks and settings for random evolution

Here anything Cell Lab related that doesn't fit into the other topics can be discussed.
Post Reply
Wilcz
Posts: 61
Joined: Mon May 14, 2018 11:17 am

Tips, tricks and settings for random evolution

Post by Wilcz » Wed Aug 01, 2018 3:22 pm

Today I got lucky and from random cells and radiation I evolved some simple swimming oeganisms. They didn't last long though. A while back I was also experimenting with mutation of floating plants. I made a simple floater with a photocyte and a buoycyte and just by random chance (and some tweaking) I ended up with a monster that used virocytes to hijack other organisms. Not that it's anything special and new but pretty amazing to watch natural selection in action.

I keep these settings constant almost all the time:
Radiation: 0.010-0.0015
Nitrates: 0.205-0.215

Other settings may vary based on what you want to achieve but I also limit nutrients and make them very lumpy just so simple phagocytes don't have it too easy. I sometimes add gravity just so they sink to the bottom and don't spread like crazy all the time. What about you?
User avatar
Petter
Site Admin
Posts: 659
Joined: Mon Dec 08, 2014 11:03 pm

Re: Tips, tricks and settings for random evolution

Post by Petter » Tue Aug 21, 2018 1:10 pm

This is an interesting topic, might warrant its own page in the wiki. I find it very useful to limit the number of cell types, perhaps some view that as cheating. A swimmer is evolved pretty quickly with just flagellocyte and phagocyte enabled. Every extra mode enabled makes it take a lot longer.
User avatar
H4yw1r3
Posts: 521
Joined: Thu Jul 13, 2017 5:19 pm
Contact:

Re: Tips, tricks and settings for random evolution

Post by H4yw1r3 » Tue Aug 21, 2018 1:47 pm

Maybe that's why 'smart' organisms can't evolve. Too much types of cells and modes are involved to evolve what the user wants.
Are complex creatures too advanced for you to understand?
Well, let me flood you with them.
User avatar
Alast
Posts: 2181
Joined: Thu Jul 02, 2015 9:15 am

Re: Tips, tricks and settings for random evolution

Post by Alast » Tue Aug 21, 2018 2:52 pm

It certainly plays some role in it, yes. So many cell types and then so many settings need to work together.
Perfection hasnt reached me yet, but its trying hard!
Wilcz
Posts: 61
Joined: Mon May 14, 2018 11:17 am

Re: Tips, tricks and settings for random evolution

Post by Wilcz » Wed Aug 22, 2018 9:59 pm

Petter wrote:
Tue Aug 21, 2018 1:10 pm
This is an interesting topic, might warrant its own page in the wiki. I find it very useful to limit the number of cell types, perhaps some view that as cheating. A swimmer is evolved pretty quickly with just flagellocyte and phagocyte enabled. Every extra mode enabled makes it take a lot longer.
I have never thought of that! Will definitely try.
H4yw1r3 wrote:
Tue Aug 21, 2018 1:47 pm
Maybe that's why 'smart' organisms can't evolve. Too much types of cells and modes are involved to evolve what the user wants.
Might be. But it's also hard to get all the signalling, sensing and reacting right with just natural selection and limited number of cells. With high enough cell numbers, dish size and some variable conditions sure, but my phone hits the limit at around 1500.
Wilcz
Posts: 61
Joined: Mon May 14, 2018 11:17 am

Re: Tips, tricks and settings for random evolution

Post by Wilcz » Fri Aug 24, 2018 12:06 pm

An observation I made today: If you run a simulation for a while with all of the cell types enabled (Gene pool) and then disable some of them they will still show up from time to time and I'm pretty sure I know why.

So let's say you have the whole gene pool enabled. You let the simulation run for a while and some modes mutate and become, let's say, nitrocytes. You disable the nitrocytes in the gene pool but after a while they start showing up anyway. The gene pool simply doesn't disable the cell types already active in modes! And that's pretty interesting because if you do that to a sensocyte (or any other cell that outputs signals) the organism has a much higher chance of making a use of it. Note that the chance is still pretty small but I feel like it's much higher than with all the cell types enabled.

So in a nutshell: modes keep their cell types even if you disable these cells in the gene pool!
User avatar
Alast
Posts: 2181
Joined: Thu Jul 02, 2015 9:15 am

Re: Tips, tricks and settings for random evolution

Post by Alast » Fri Aug 24, 2018 12:45 pm

Which makes sense, they're just not available to the mutation pool anymore :)
Perfection hasnt reached me yet, but its trying hard!
Wilcz
Posts: 61
Joined: Mon May 14, 2018 11:17 am

Re: Tips, tricks and settings for random evolution

Post by Wilcz » Wed Sep 05, 2018 3:52 pm

Another observation of the same substrate.

I had organisms that anchored themselves (simple gluocyte+photocyte colonies) to stay as close to light but after a while they evolved into a simpler form (the organism became mobile and consisted only of 2 photocytes). However, after reducing light range the organisms started anchoring themselves to the walls again. The species basically devolved into the previous form using modes its ancestors used.
Post Reply