Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[question] Timestep unit in Simulation #56

Open
undulating-moon opened this issue Sep 11, 2022 · 2 comments
Open

[question] Timestep unit in Simulation #56

undulating-moon opened this issue Sep 11, 2022 · 2 comments

Comments

@undulating-moon
Copy link

According to Polychrom documentation, the timestep of simulation is always fs scale (integrator="langevin"). But I need simulation in ms scale to fit experiment data, just setting timestep=1e12 fs will lead to the following error:
polychrom.simulation.EKExceedsError: Ek=4.425593696827154e+29 exceeds 10

Is there any method to solve this problem? Will just normalize time units work?

Thanks in advance for your guiding.

@mimakaev
Copy link
Collaborator

Our timestep does not represent realistic time (our bond is 1nm, and particles are 100 AMU, neither of them is realistic) and this is done partially on purpose. Overall, we don't have a way to simulate a particular amount time - it would depend on your monomer mass, size, diffusion coefficient, visco-elastic properties of the system, etc.

We recommend running a simulation for a given amount of time (in many replicates), calculating average MSD(t), and trying to fit it to the biologically relevant MSD measured in some experiment (whatever you can find for your cells/species). This way you can link your simulated MSD(t) to experimental MSD(t) and convert simulated time to experimental time.

As for the timestep - it is a technical parameter that affects langevin dynamics simulations. For all practical purposes it is safe to keep integrator as variableLangevin, and basically measure time in timesteps. Then after applying the procedure described above you can find what real time does it correspond to.

@undulating-moon
Copy link
Author

Thank you for your advices! I will try it soon, but is there any detail about fitting simulated MSD'(t') to experiment MSD(t)? Just add two parameter a,b, (MSD'=aMSD, t'=bt) and fit parameters?
Well, we have MSD(t) data generated from experiment data, and we want to use this MSD(t) data to explore force fields of polymers (DNA). I just wonder whether adding 2 parameter (a,b) will interfere the parameters of force fields.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants