Pheromone trails

Here anything Cell Lab related that doesn't fit into the other topics can be discussed.
Post Reply
User avatar
wapcaplet
Posts: 363
Joined: Thu Dec 08, 2016 9:52 pm

Pheromone trails

Post by wapcaplet » Tue Dec 20, 2016 7:41 pm

I've read about how ants will deposit pheromones to create a trail to a food source, and have been wondering if such behavior could work in Cell Lab. I have not tried any experiments yet, but here's one approach I'm thinking might work:
  • Substrate has zero nutrient density, but lots of light. Photocyte-based organisms produce the primary food source.
  • There's a central colony/queen/hive where new ants are born and nourished
  • Ants spread out from the colony, seeking the photocytes/plants to get food
  • Those that find food eat their fill, then work their way back to the colony. Along the way, they deposit "pheromones" in the form of small cells that die and leave a bit of nutrient
  • Ants that make it back to the colony deposit their collected nutrients in lipocytes (honey) or into the queen so she can birth new ants
  • The trail of nutrients back to the colony is followed by subsequent ants, giving them just enough nutrient to keep moving, and leading them to the "real" food
I guess the trail could also be secrocytes, but it can't just be "food smell" or else they'll get distracted. Seems like actual nutrients would give the best chance of survival on the way there and back.

There's a lot of complex behavior here I'm not sure how to accomplish--unless the food is close to the colony, ants may wander at random and never find food, or never find their way home. The "hungry" ants would need to prioritize following trails and finding photocytes, while the "full" ants would need to make getting home their top priority. From what little I understand of real ants, the advantage of the pheromone trail is when it's reinforced over time by many other ants following it. When the food source dries up, the ants won't deposit more pheromones, allowing the obsolete trail to dissipate.

Has anyone tried anything like this? Think it could work?
User avatar
Nayus
Posts: 679
Joined: Thu Oct 06, 2016 2:03 am
Location: Argentina

Re: Pheromone trails

Post by Nayus » Tue Dec 20, 2016 10:09 pm

It sounds really complicated... But I don't think it'd be impossible :) . Do you plan to try it or want to form some kind of team for the problem? :lol:
User avatar
wapcaplet
Posts: 363
Joined: Thu Dec 08, 2016 9:52 pm

Re: Pheromone trails

Post by wapcaplet » Wed Dec 21, 2016 3:57 am

Ha! My skills aren't up to the task yet, mostly I am just looking for opinions from more experienced players like yourself about whether it could be done, or ideas for how. I was particularly inspired by your drone/queen creations (and it still blows my mind that you did all that with just one genome). Since a substrate can have many different genomes on it, I don't have any expectation of being able to do it all with just 20 modes--the ants, queens, and food sources could certainly be separate species.

One of the first sub-challenges for something like this is a critter that can do a kind of decision-making or prioritization, so it could have two or three different goals, "collect food", "feed queen", "follow trail". Stemocytes will certainly help, but I also wonder if the simple mechanism of cell splitting could accomplish part of it as well.

For example, assume there are two different stereocyte modes, say M5 (sensing food) and M6 (sensing yellow cells), and the ant starts with just M5, so it would look for food. When the ant gets full, M5 splits and is replaced by M6. So it would go from food-seeking to yellow-cell-seeking (ignoring food). Maybe yellow cells are the queen; this would (in theory) make the ant want to go feed the queen only when it has enough food for her. I'm not sure how it could switch from M6 back to M5, though, so this would be a one-way trip.

Or maybe splitting could be combined with stemocytes, so when some goal is reached (say, stemocyte detects S3), it switches modes, pops off unwanted stereocytes/sensocytes, and replaces them with different ones, plus a new stemocyte that can switch it back again later.

Just as a proof of concept of the alternating stemocyte idea, I made this substrate with some red and blue cells (with attached S3/S4 secrocytes), and a swimmer that has a stemocyte "indicator light" of the last color it touched. There's no behavioral change; it's just a blinky light :-)
wapcaplet-red-blue-switchers.substrate

I will keep tinkering, but would welcome any suggestions or contributions.
User avatar
Nayus
Posts: 679
Joined: Thu Oct 06, 2016 2:03 am
Location: Argentina

Re: Pheromone trails

Post by Nayus » Wed Dec 21, 2016 4:53 am

:idea: I think that maybe the ants could have a devorocyte. And the 'trail' actually be Photocytes that don't split. The Ants when they have enough food start looking for the color of the Photocytes instead of the food. They could follow their own trail (or others) back to the Queen.

So the ants would have a early life where they lay those little photocytes that grow to be the trail. Then the late life they change its stereocyte.

The difficult thing is to make the trail, I think.

My Ant Queen was just one genome because we can't make a organism reproduce an egg of other organism. So I couldn't have the queen and the ants have different genomes 'cause I just coulnd't make it reproduce from one to the other.

The only way we have currently, I think, of contact between organisms are the examples of Sexual reproduction by Secrocytes or Virocytes. Those cells are the only ones that produce some signal 'outside' it's own organism and can trigger a change in other genome. I think a glueocyte and a intense system with a neurocyte could do that too but anyway those are the only ways.
User avatar
wapcaplet
Posts: 363
Joined: Thu Dec 08, 2016 9:52 pm

Re: Pheromone trails

Post by wapcaplet » Wed Dec 21, 2016 5:51 am

Nayus wrote:I think that maybe the ants could have a devorocyte. And the 'trail' actually be Photocytes that don't split.
That would definitely make the trail more permanent, I guess as long as viscosity or friction is enough to keep them from getting nudged around by passing ants. With a trail so rich in nutrients, I wonder if the ants would even need to travel all the way to the "real" food before getting full. On the other hand, the tiny nutrient left by a dead cell won't move, but dissipates very quickly (only lasting about 6 or 7 seconds).
Nayus wrote:My Ant Queen was just one genome because we can't make a organism reproduce an egg of other organism.
Ah, of course; I did not even think of that. And virocytes can only copy modes, so they wouldn't help to gain extra modes. I suppose the queen can be relatively simple--she doesn't need complex behavior or mobility, but is just a repository of nutrients to create new eggs.

Perhaps instead of a single species of ant, if 20 modes is not enough, there could be two cooperative species--say, one to lay the trail, and another to gather food--and the egg-laying parts of both species could be glueocyted together into an uber-queen that stores nutrients and lays eggs for both of them. Ooh, there are all sorts of directions that could go in...
User avatar
bwisialo
Posts: 2953
Joined: Sat Aug 15, 2015 10:22 pm

Re: Pheromone trails

Post by bwisialo » Fri Dec 23, 2016 1:27 am

I tried the "trail of bread crumbs" stategy to get pack hunting predators. They alternately produce colored cells and new predators. They have two Senseocytes, one for prey, one for the colored cells. It was almost but not quite working. Maybe I just need to try harder or change the substrate or prey a bit.

It might work with drones and queen better. One difficulty might be that drones / ants could get onto a trail, but not "know" which direction goes toward and which goes away from the queen, or there might be so many "bread crumbs" around that the drones don't follow a trail, per se, but just sort of wander around eating them and then going toward a queen. You'd probably need life stages where they eat in one stage and then go to the quuen in the other. And they couldn't do multiple trips out and back, but that's a whole other ball game.
amor fati
User avatar
wapcaplet
Posts: 363
Joined: Thu Dec 08, 2016 9:52 pm

Re: Pheromone trails

Post by wapcaplet » Sun Dec 25, 2016 9:16 pm

bwisialo wrote:And they couldn't do multiple trips out and back, but that's a whole other ball game
Right, the more I think about it, that seems impractical. In quantity, one trip each would be plenty. A viable single round-trip organism would need two "behavior modes", or life-cycle phases or whatever we want to call it, with two different dominant priorities "eat plants (fill up with nutrients), feed queen (dump nutrients out and die)".

I've been playing around with a swimmer that switches between seeking red and blue cells. My red cells emit S3. When a stemocyte gets S3, it switches/splits into itself, plus a new stereocyte that feeds a strong blue-cell signal to the flagellocytes. When the same stempocyte gets S4, it switches/splits into itself, plus a strong red-cell-detecting stereocyte (popping off the blue-detecting stereocyte). Repeat ad infinitum. It kinda works, though they don't actually consume the red/blue cells, only bump into them, and leave red/blue stereocyte droppings everywhere.

But now I'm leaning more toward using full lipocytes as the trigger for the behavioral change. Lipocyte split mass and ratio have many curious effects on adhesin and the overall organism shape, and I've seen even seen them (apparently) consume a neighboring cell while splitting. Since goal #1 is "fill up with nutrients", lipocyte split is the perfect time to switch stereocytes and behavioral priorities.

What about using S3 or S4 (for example) itself as the pheromone? I just thought of this and haven't tried it yet, but what if ants drop an S3-emitter on a food source, and other ants can smell S3. Would a big pile of S3 cells increase the concentration enough for them to smell it further away?
User avatar
bwisialo
Posts: 2953
Joined: Sat Aug 15, 2015 10:22 pm

Re: Pheromone trails

Post by bwisialo » Mon Dec 26, 2016 6:45 pm

Smelling signal substance is not something Stereocytes can do.
amor fati
User avatar
CandyYAHFT
Posts: 780
Joined: Fri Oct 23, 2015 7:10 pm

Re: Pheromone trails

Post by CandyYAHFT » Fri Dec 30, 2016 7:56 am

The current way in which substances work is that from the moment they are released. They quickly get to zero, if not constantly produced. While this is true with some real life substances. More accurate and complex behaviour could be done if we had a way to implement lingering effects to substances. As a way to leave a trail.
PEACE & LOVE,
CANDY
User avatar
bwisialo
Posts: 2953
Joined: Sat Aug 15, 2015 10:22 pm

Re: Pheromone trails

Post by bwisialo » Fri Dec 30, 2016 9:06 am

The time that signal substance lasts doesn't matter for the desired effect here, as I understand it. Cells that produce substance release substance into themselves and connected cell, or into cells they contact in the case of Secrocytes. They don't release signal substance onto the substrate.
amor fati
Post Reply