Hi eddie
Recently I used charmm forcefield to simulate surface tension of liquids. I
generated inputs based on charmm gui (I did my simulations with gromacs).
Although charmm forcefield stipulates that dispersion corrections should
not be used I noticed that when I exclude dispersion corrections the
surface tensions were significantly lower compared to experimental data. I
came to the conclusion that dispersion corrections are indeed needed and
that dispersion correction to the pressure tensor is not negligible.
Whether to use dispersion correction or not then becomes a choice depending
on what property you want to use
Will excluding dispersion correction adversely affect your simulation
results ? (self assembly, surface tension)
Then I believe you have to include dispersion correction and state that in
the manuscript/article/report that you write.
MD engines typically provide two options to account for this, force shit
and energy shift.
I beleive force shift has not yet been implemented in gromacs but LAMMPS
has this option.
Since accounting for tail corrections is force field specific it would not
be mentioned in an MD engine manual. As a courtesy gromacs developers have
mentioned on how to acount for vdw corrections for the charmm force field.
Hope this helps
Cheers
Arun
On Wed, May 31, 2023 at 5:40 PM Eduard Neu <eduard.neu.fau.de> wrote:
> Hi Arun,
>
> thanks for your answer!
>
> I was wondering why one needs to apply dispersion corrections specifically
> when using the AMBER force field.
>
> For example, using the CHARMM force field the dispersion correction is not
> needed. See here:
> https://manual.gromacs.org/current/user-guide/force-fields.html
>
> Also, it was brought to my attention that AMBER by default sets vdwmeth=1
> (p.408 AMBER22 manual). This means that by default a continuum model
> correction for energy and pressure is used. I assume this is also the
> default setting for the AMBER force field.
>
> Additionally, this is explicitly stated in the ff14SB publication:
>
> "A continuum model correction for energy and pressure was applied to
> long-range van der Waals interactions."
>
> However, I couldn’t find any information on using
> vdw-modifier=Potential-shift. I guess this is GROAMCS specific:
>
> https://manual.gromacs.org/current/reference-manual/functions/nonbonded-interactions.html
>
> Best
> Eddie
>
>
> On 5/26/2023 7:47 PM, Arun Srikanth wrote:
>
> Hi Eddie
>
> You need to apply dispersion corrections for energy and pressure. For
> pressure the correction cannot be neglected
> More on this you can find here
>
> https://manual.gromacs.org/current/reference-manual/functions/long-range-vdw.html
> Arun
>
> On Fri, May 26, 2023 at 7:05 PM Eduard Neu via AMBER <amber.ambermd.org>
> wrote:
>
>> Hi everyone,
>>
>> I also posted this question in the GROAMCS forum, but haven’t received a
>> satisfying answer yet:
>>
>>
>> https://gromacs.bioexcel.eu/t/mdp-parameters-for-non-bonded-interaction-using-the-amber-force-field/6458
>>
>> In the GROMACS manual there are no clear recommendation regarding the
>> non-bonded parameters when using the AMBER force field:
>>
>>
>> https://gromacs.bioexcel.eu/t/mdp-parameters-for-non-bonded-interaction-using-the-amber-force-field/6458
>>
>> I came across this post:
>>
>>
>> https://gromacs.bioexcel.eu/t/protein-ligand-complex-amber-production-question/1247/9
>>
>> in which following parameters are recommended:
>>
>> constraints = h-bonds
>> rcoulomb = 1.0
>> rvdw = 1.0
>> vdw-modifier = Potential-shift-Verlet
>> DispCorr = EnerPres
>> coulombtype = PME
>> fourierspacing = 0.125
>>
>> Based on how ff19SB and lipid21 were tested a cutoff of 10 Åis reasonable.
>>
>> However, I don’t understand why vdw-modifier = Potential-shift-Verlet
>> and DisCorr = EnerPress is used.
>>
>> Is there any literature on this available or am I missing something out
>> in the original publications of the force fields?
>>
>> Additionally, CHARMM-GUI generated outputs that use vdw-modifier = None.
>>
>> Regard
>> Eddie
>> _______________________________________________
>> AMBER mailing list
>> AMBER.ambermd.org
>> http://lists.ambermd.org/mailman/listinfo/amber
>>
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed May 31 2023 - 10:30:03 PDT
