Comparing all-atom and coarse-grained simulation

GROMACS version: gromacs/2019.4
GROMACS modification: No

Hi everyone,

I’m comparing all-atom (AA) and coarse-grained (CG) simulation of a bilayer membrane containing 72 POPC lipid molecules. CHARMM36 force field was used for AA simulation, MARTINI force field was used for CG simulation, and all parameter settings are the same between the 2 systems to be run on GROMACS. After energy minimization and 12 ns of equilibration (4 ns NVT and 8 ns NPT) were run, the temperature and pressure were stable at 310K and 1 atm. Then the membrane was deformed under equibiaxial stretching, with deformation velocity is applied in x and y directions. After that, the von Mises stress-strain curve was plot (stress in y axis and strain in x axis).

  1. Comparing the two curves, the stress of CG system over time is much lower than that of AA system - about half the values. I’m not sure why this happens.

  2. Besides MARTINI, I also try united atom force field C36UAr, after deformation the stress values for this system is similar with AA system. However, the average temperature after 12ns of equilibration is about 312K, which is 2K higher than the preset 310K values.

So my questions are,

  • Why is the stress in CG system lower than that in AA system?

  • Why is there 2K higher in the united atom system compared to AA system after equilibration?

If you have any idea about these questions, please give me some advice or suggest me some explanations for them. Please let me know if you need other information.

Thank you very much and have a great day.

Hi Amy,

I hope, more people with CG experience chip in on this one.

The first explanation that comes to my mind is that there is less friction between particles in the coarse-grained system, because there is much less of entropic contributions due to the reduction of multiple atoms into single beats. Especially during some non-equilibrium process like the deformation that you perform, I would expect that you would have to exert much more work on the MD system than the CG system to reach the same curvature.

1 Like

Hi cblau,

Thank you very much for your answer! Could you please explain a bit more about how less friction or less entropic contribution leads to (or related to) lower stress value?