Creating an amyloid fibril spanning infinitely in the z-direction using pbc

GROMACS version:2020.4
GROMACS modification: Yes/No
Here post your question
Good day everyone!
I want to simulate an amyloid fibril spanning infinitely in the z direction and is able to interact with the periodic molecules only in the z direction. I have a few questions regarding this
Q 1. I turned on
periodic-molecules= yes and pbc = xy ( as described in the manual for infinite z systems);

but pbc = xy wouldn’t work. Some issue came up with the verlet cut-off scheme. and insists on nwalls=2 so that pcoupl and ewald sums would work. So, that won’t be good.

So as i understood from a recent work, (link below), i reverted back to pbc=xyz and gave more space in the x and y dimensions so that there won’t be any interactions in x and y dimensions. This is what i understood and is the only possibility. Is my approach correct?

Q 2: regarding box creation through editconf. the length of my fibril in z axis is 2.9 nm. Inorder to create a periodic system along z how should i create a box?
a) a box shorter than 2.9 nm say 2.5 nm? or
b) a box exactly at 2.9 nm ?
c) a box with 2.9 nm + 0.49 nm ? (corresponding to hydrogen bonds)

Comments/criticism would be valuable.
thank you.!

This option is only needed for molecules whose bonds cross periodic boundaries. You should not use it for this system.

This is correct.

If you have specific questions about a published paper, you should contact the authors directly to ask for example inputs.

1 Like

thank you , i will do as you suggested.!

hi again,

I wanted to try what i can till i get a response,
So i got a couple of follow up questions

Professor, as soon as i remove the restraints for the production MD, the fibril drops out of the box; So how can i perform the simulation without using
periodic-box=yes? ;
it doesn’t move much only when it is such a case

  1. My fibril has just three protofibril units, which is very small along the z axis, so cut-off errors come up with box parameters. So i used pymol to create 5 protofibril units stacked along the z axis and created independent chain labels and rebuilt it; but during the simulation, a couple of units behave independently. Am i missing something after creation of the fibrils?

  2. If there is any protocol for creating infinite fibrils for gromacs, could anyone please share it;


I don’t understand your descriptions of points 2 and 3. Please upload images of what you are observing. If the chains are not remaining in alignment, then you haven’t properly oriented them and set the box size.

I have made it available at this link, Professor…
with permissions to; I hope this is fine;
thank you!

Your attempt to translate a copy of the peptide has led to misalignment. The original coordinates need to be aligned such that the hydrogen bonds are directly coincident with one coordinate axis, and then other copies of peptides can be translated along this axis.

1 Like

Thank you Professor, for your insights…
It works.!
I made a mistake in calculating the dimensions of the fibril using a pymol script, which didn’t account for the non planarity of the fibrils. So, all is well and thank you.!