[lammps-users] 14 positions on the cylinderical surface connected with 14 beads of a bead spring model

Dear LAMMPS developers,

I am trying to do Langevin Dynamics simulation of a chromatin model(DNA as a bead spring model + protein as a cylinder).
As a first step, I would like to choose 14 positions on the surface of the cylinder (protein) and these 14 positions are connected
to the 14 beads of the DNA bead spring model through Morse potential. I am new to LAMMPS and learning this simulation
step by step. I have attached the LAMMPS code that I am writing and its not complete at this stage. I want help to model the

polymer1.lam (7.01 KB)

your input is full of commented out and contradicting commands that override each other and thus are difficult to make sense of.
also, you will be hard pressed to find somebody that takes your input and completes it for you.
that is particularly difficult without knowing the specific details of your test system.

you have much better chances to get help from the mailing list if you ask more specific questions, i.e. describe exactly where you are stuck and what you have tried to realize your model and where it failed and how.

also, you will be more likely to get help, if you “clean up” the inputs that you post from all superfluous content, commented out code and comments for your own reference.
in many cases it would be helpful to not post your original research project, but a simplified minimal example.

some more general recommendations for posting to the lammps-users mailing list are here:


Dear Axel,

Thank you for your feedback and now, I have completely removed the superfluous information.

Model : DNA bead-spring chain is wound around the protein cylinder.

Aim : Unwind the DNA from the protein

I am trying to do Langevin dynamics simulation to unwind the DNA from the protein (chromatin). For this,

I have modeled the DNA as a bead spring chain with periodic boundary using the atomic style full and computed the mean squared displacement
and the energy.

My question is, how one can model the protein as a cylinder in LAMMPS. If so, how to choose 14 positions on the cylinder
and these 14 positions will be connected via Morse potential to the DNA bead-spring chain model having 14 beads.

From the documentation I understood one can use the following commands for cylinder and Morse potential:

**region void cylinder y 2 3 5 -5.0 EDGE units box**

**pair_style     morse 2.5                ** 
**pair_coeff 1 1 100.0 2.0 1.5 1.0** 

I am stuck up in choosing the 14 positions on the cylinder and connecting these positions with DNA beads through Morse potential. 
What is the correct way to implement this in LAMMPS?.



Women Scientist Fellow


polymer3.lam (1.1 KB)

from what I gather you have two separate problems:

  • choose a suitable model for your simulation (this is only marginally relevant for this mailing list)
  • which commands to use to implement that model (this is on-topic for this mailing list)

but you seem to be mixing those two and thus are making yourself confused.

it is rather pointless to discuss how to implement something unless you are clear about the model and that doesn’t seem to be resolved yet.
you say you want to model your proteins “as a cylinder” but you don’t say how the bead-spring chains should be interacting with it.
you need to decide on that first before you can discuss implementing it.
please note that a “region” in LAMMPS is just a region and has no interactions with anything per se. it may be used as bases for a wall potential, though or to define a region to be filled with atoms or where to delete atoms or to define a group of atoms and so on. more details are in the manual.

as for how to attach your chains to points in space: you can just use additional atoms that are immobile (i.e. not included in the time integration group(s)) and have no interactions except a bond to the corresponding tether atoms of your chains.

some comment about your input. you have:

pair_style soft 1.12246152962189
pair_coeff * * 100.0 1.12246152962189

pair_style morse 2.5
pair_coeff 1 1 100.0 2.0 1.5 1.0

the second pair_style command will wipe out the first. there can be only one pair style at a time. if you want different functional forms for different interactions between atom types, then you can use pair style hybrid.

please also note that atom style “full” is overkill for your system since you don’t have any charges or dihedral styles or improper styles defined.