Influence molecular dynamics parameters on membrane model parameters

User discussions
1

GROMACS version: 5.0.4 (running md), 2018.4 (analyze parameters)
GROMACS modification: No
I try to test some new POPC parameters (charges and dihedrals), so I build and run simulations. I build 10x10 POPC membrane and compare area per lipid, the width of membrane and molecular order parameter to these from the experiment. I use OPLS-AA force field. This is one of my .mdp file
First:
title = MD equilibration;
tinit = 0
cpp = /lib/cpp
constraints = hbonds
constraint_algorithm = lincs
integrator = md
dt = 0.002 ; 1 fs
nsteps = 5000000 ; 10 000 ps
nstcomm = 1
comm_mode = linear
; output
nstxout = 5000
nstvout = 5000
nstxtcout = 500
nstfout = 0
nstlog = 500
nstenergy = 500
cutoff-scheme = Verlet
;
ns_type = grid
nstlist = 10
pbc = xyz
coulombtype = PME
rvdw = 1.0 ; 1.8
rlist = 1.0
rcoulomb = 1.0
fourierspacing = 0.1
pme_order = 6
ewald_rtol = 1e-6
; temperature coupling
tcoupl = berendsen
tc-grps = water non-water
tau_t = 0.5 0.5 ;
ref_t = 300 300
; pressure
Pcoupl = berendsen
Pcoupltype = semiisotropic
tau_p = 5.0
compressibility = 4.5e-5 4.5e-5 4.5e-5 0.0 0.0 0.0
ref_p = 1 1 1 0.0 0.0 0.0
gen_vel = yes
gen_temp = 300
gen_seed = 730624

Second:
title = MD equilibration;
tinit = 0
cpp = /lib/cpp
constraints = hbonds
constraint_algorithm = lincs ;
integrator = md
dt = 0.002
nsteps = 150000000
nstcomm = 1
comm_mode = linear
continuation = yes
; output
nstxout = 5000
nstvout = 5000
nstxtcout = 500
nstfout = 0
nstlog = 500
nstenergy = 500
cutoff-scheme = Verlet
;
ns_type = grid
nstlist = 10
pbc = xyz ; no
coulombtype = PME
rvdw = 1.0 ; 1.8
rlist = 1.0
rcoulomb = 1.0
fourierspacing = 0.1
pme_order = 6
ewald_rtol = 1e-6
; temperature coupling
tcoupl = nose-hoover
tc-grps = water non-water
tau_t = 0.5 0.5 ;
ref_t = 300 300
; pressure
Pcoupl = parrinello-rahman
Pcoupltype = semiisotropic
tau_p = 5.0
compressibility = 4.5e-5 4.5e-5 4.5e-5 0.0 0.0 0.0
ref_p = 1 1 1 0.0 0.0 0.0
;
gen_vel = no

I try to use parameters from one of the article, but I forget to add option
DispCorr = EnerPres
and I am curious and influence of my membrane parameters (I will check them later).

So I have questions (maybe you know answer for even one of them, even small information would be helpful)

  1. What do you think about my mdp files? What I can change to have better results?
  2. Is adding DispCorr = EnerPres will have a big influence on parameters like area per lipid, Smol, etc.
  3. I think about constraints = hbonds . I see that some people use constraints = all-bonds, which constraints I should use in which simulation?
  4. I use Pcoupltype = semiisotropic , but I also time to time use anisotropic, when my simulation is not too long, so my box not deformed too much. My simulation run only for 300 ns, so I could possibly use anisotropic. Isotropic, I use for example when I have protein in water or something like that.
  5. These parameters I took from the previous publication.
    fourierspacing = 0.1
    pme_order = 6
    ewald_rtol = 1e-6
  6. What could happen if I use comm-mode Angular instead of Linear (I know that Linear is only for condensed-phase systems, angular is only for vacuum simulations, but I don’t know why)?
  7. Do you know any other important parameters that I can check for membranes? I think about lateral diffusion, but it’s hard to find experimental value? Now I have only three: area per lipid, a width of the membrane, and molecular order parameter.
  8. Do you think that using a newer version of Gromacs can have the influence of membrane parameters? I use 5.0.4 (in the article they use 4.6.5)?

Thanks in advance. Even small information will be useful.