Mdp parameters for non-bonded interaction using the AMBER force field

GROMACS version: 2023.0
GROMACS modification: No

Hey,

I was wondering if there is a clear recommendation for the mdp parameters regarding the non-bonded interactions using the AMBER force field?

The GROAMCS manual recommends following parameters when using the CHARMM force field:

constraints = h-bonds

cutoff-scheme = Verlet

vdwtype = cutoff

vdw-modifier = force-switch

rlist = 1.2

rvdw = 1.2

rvdw-switch = 1.0

coulombtype = PME

rcoulomb = 1.2

DispCorr = no

However, there is no such recommendation in the manual for the AMBER force field.

Doing some research, I came across this post and this post from @alevilla .

In these posts following parameters are recommended for the AMBER force field:

constraints = h-bonds ; bonds involving H are constrained

rcoulomb = 1.0 ; short-range electrostatic cutoff (in nm)

rvdw = 1.0 ; short-range van der Waals cutoff (in nm)

vdw-modifier = Potential-shift-Verlet ; Amber specific

DispCorr = EnerPres ; account for cut-off vdW scheme

coulombtype = PME ; Particle Mesh Ewald for long-range electrostatics

fourierspacing = 0.125 ; grid spacing for FFT

Doing some further research a came across this website that explained how to choose the appropriate cutoff distance. For the AMBER force field, it is 10 Å or 1 nm. The cutoff value depends on the cutoff which was used during the parametrization of the force field. As stated in the primary literature ff99SB, ff14SB, ff19SB and lipids21 were all parameterized using a cutoff value of 10 Å or 1 nm, respectively.

However, reading the primary literature I couldn’t find a reason why

vdw-modifier= Potential-shift
DispCorr = EnerPres 

are used or why specifically these values are chosen.

I also checked the Sorin Lab website for ffAMBER but I couldn’t find an explanation there.

What is the reasoning for using

vdw-modifier = Potential-shift

DispCorr = EnePress

when using the AMBER force field in GROMACS and is there any literature or reference for this available?

Also, if these are the general non-bonded parameters in the mdp file when using the AMBER force field, wouldn’t it be useful to include them in the force field section of the GROMACS manual?

Additinoally, I also use

coulomb-modifier = Potential-shift

when using the AMBER force field. Again in there any literature or reference availbe for this parameter using the AMBER force field?

Cheers
Eddie

Hey,

In case you are working with ff14SB, I think that one uses 8 Å cutoff for non-bonded interactions:
“Nonbonded interactions were calculated directly up to 8 Å.” from the ff14SB publication.

Please correct me if I am missing something.

Regards
Daniel

Hey Daniel,

thank you very much for this advice! You are right the cutoff for non-bonded interactions is 8 Å in ff14SB. I double cheeked the information for each force field and these are the used cutoffs.

ff99SB 8 Å:
„A cutoff of 8-Å was used for nonbonded interactions, and long-range electrostatic interactions were treated with the particle mesh Ewald (PME) method“

ff99SB-ILDN 10 Å:
„A cutoff of 10 Å was used for the Lennard-Jones interaction and the short-range electrostatic interactions.“

ff14SB 8 Å:
„ Nonbonded interactions were calculated directly up to 8 Å.“

ff19SB 10 Å:
„The direct space nonbonded interaction cutoff was 10.0 Å for explicit solvent simulations and 9999.0 (no cutoff) for implicit solvent simulation.“

However, there are separate force fields for lipids in the AMBER force field.

Lipid14 10 Å:
„PME was used to treat all electrostatic interactions with a real space cutoff of 10 Å. A long-range analytical dispersion correction was applied to the energy and pressure.“

Lipid21 10 Å:
„All restraints were removed and the PME was used to treat all electrostatic interactions with a real space cutoff of 10 Å“

In the current Amber manual (p. 51) it is stated that the behavior of lipids using the AMBER force field is highly sensitive to simulation conditions and a cutoff of 10 Å is recommended:

„A word of caution regarding barostat and cut off selection with lipid simulations. It is well known that lipids can be very sensitive to simulation conditions. Previous advice has been to use the Berendsen barostat and a 10A cutoff when simulating lipids with the AMBER Lipid force fields. This was due to bilayer deformation that was regularly seen with simulations run using the Monte Carlo Barostat. Recent work by Gomez et al.[91] investigated this behavior and determined that there are issues when using the MC barostat with hard Lennard-Jones cutoffs. The issue affects all simulations but is most obvious when simulating lipid bilayers. Gomez et al recommend use of an LJ force switch when running simulations with the MC barostat.“

Based on that I would use a cutoff of 10 Å as soon as there is a membrane system.

Still I couldn’t find the information that recommends the use of

vdw-modifier = Potential-shift 
DispCorr = EnePress

Additionally, I tried CHARMM-GUI to generate the inputs and these are the parameter in the mdp file:

cutoff-scheme           = Verlet
nstlist                 = 20
rlist                   = 0.9
vdwtype                 = Cut-off
vdw-modifier            = None
DispCorr                = EnerPres
rvdw                    = 0.9
coulombtype             = PME
rcoulomb                = 0.9

I would be very interested in the most suited vdw-modifier for the AMBER force field.

Best
Eddie