Fix and displace ends of filaments in MEDYAN and Cytosim

[blairlyons]

Use Case

As a developer running comparable models for the actin filament physics figure, I need to determine if there is a way to fix and displace ends of actin filaments in MEDYAN and Cytosim.

Acceptance Criteria

For MEDYAN and Cytosim:

• is there a way to fix ends of filaments?
• is there a controlled way to displace ends of filaments?
• If so, how? which objects in the API can we use?

Details

maybe we can use hand binders or microtubule organizing centers?

[mattakamatsu]

@blairlyons I haven't figured out how to actively move an end of a fiber during a simulation in cytosim, but I see how to set and fix the ends of a fiber at the beginning of the simulation. And then we could run loops of multiple simulations.

e.g.

set single pivot
{
hand = strong_hand
stiffness = 1000
activity = fixed
}

new filament
{
attach1 = pivot, 0, minus_end
attach2 = pivot, 0, plus_end
}

So, we would run a simulation with fixed singles that hold both ends in place, and see how the filament relaxes. Use the coordinates at the end of the simulation to run a new simulation, with one end positioned closer by a small amount and fixed in place. And then repeat. So we'd use vivarium to read in the filament positions at the end of each simulation and move e.g. the plus end one step closer in X every simulation. This would also allow us to switch between ReaDDy and Cytosim (e.g. run a simulation in ReaDDy while pushing one end toward the other, and then switch to cytosim with the end moved a little closer, simulate, switch back, etc).

What do you think?

[blairlyons]

ok! running serial simulations should be easier now that we have a scan function in vivarium. sounds good!

@ritvikvasan do you know if there's a way to do this in medyan?

[ritvikvasan]

After talking to Aravind, it sounds like we want to use the PINBOUNDARY special protocol to pin one end of the filament. To move one end of the filament via a displacement boundary condition, Aravind suggested either using the AFM special protocol (this is not mentioned in the online docs and I don't see it in my code version. Aravind said he will reach out to the devs to ask if this is updated in the latest version). The AFM protocol may be a little hacky because the way it works is that it has some sticky actin filaments on it that eventually connect to loose actin filaments via some kind of cross-linker interaction. In our case, we would want to set the stiffness of that to be infinite so it doesn't affect the mechanics. An alternative to this would be to add our own special protocol that displaces one end of the filament by a specified amount