simtk.org

Hello all,

I have a polymer with each repeating unit as a peptide sequence. 15 units in total.
I setup the simulation from a stretched out position, I set my box size to +2nm buffer on all sides. But the issue is that if the polymer rotates, it will interact with its periodic image because the height dimension of the box is much smaller than the length.
I know a possible way to avoid this is to set all dimensions of the box to the stretched out length with buffer but that makes the simulation computationally very expensive (~ 800k atoms).

Is there a way to ensure that the polymer does not rotate w.r.t to the box (or in another frame of reference the box rotates with the system) without introducing a biasing force?

Best,
Hassan

Without introducing a biasing force? No. Anything that prevents it from rotating is, by definition, a bias.

A biasing force would be a fairly standard way of handling this situation. Yes, it would affect the results compared to letting it freely rotate in the simulation box, but hopefully the difference would be very small. It depends on what you want to learn from the simulation, of course.

Another question is whether you really need to simulate a fully stretched out chain. Are you particularly interested in that case, or was it just a convenient starting point? If you expect the chain to collapse and spend most of its time in a more compact state, you should be able to handle the collapsed chain with a much smaller box.

So I'm assuming the constraint would have to be implemented by fixing a dihedral somewhere along the polymer backbone?

The chain does collapse eventually into a more compact structure but despite that it is longer in the length dimension, and the ultimate objective is to run random mutants therefore I seek a more generic way to make sure that the image-image interaction does not occur.