[gmx-users] Protein-ligand system simulations with GROMACS

Dries Van Rompaey dries.vanrompaey at gmail.com
Wed Feb 24 20:27:44 CET 2016

Hi Guillem,

Generally, it's advised to have a solute-box distance on both sides at
least equal to your longest cutoff. This makes sure that you don't
experience artefacts due to periodic images. In my opinion, it's better to
be on the safe side here - a bigger box allows for slight unfolding events
or box shrinkage due to NPT pressure regulation, where a smaller box might
not. This can lead to unphysical interactions between copies of your
protein. Also, try not to mix 'cutoffs' and 'solute-box-distance', they're
two quite different things.

You're right that the dodecahedron does't appear as such in some
visualisation programs. This will indeed be fixed if you use trjconv with
the -ur compact flag (use your ions tpr file, that's just one extra step).
I'm not sure if you can do this in a more efficient way - I'd be happy to
learn about this myself.

Your protein will move around, that's pretty much a given. However, you
can't describe those motion as going 'outside' of the box, since you're
working in a periodic system. A lot of useful info can be found here:

Kind regards


On 24 February 2016 at 14:13, Guillem Prats Ejarque <
Guillem.Prats.Ejarque at uab.cat> wrote:

> Dear Dries,
> First of all, I performed my trials in basis of the protein-ligand
> tutorial that you say. However, it seemed to me that there was too much
> waters and I tried to reduce the box. I didn't thought with the periodic
> image convention. A cutoff of 0.9 would be ok?
> When I try to center (using the -c flag) in a dodecahedron box, it appears
> in a vertex of the "dodecahedron" (it not seems a dodecahedron), as it not
> happens in a cubic box. I have read that this is due I'm  working with a
> triclinic cell, and that I should to arrange it with trjconv -ur, but it
> requests to me a .tpr file that I don't have prior the addition of the ions.
> Another strange thing that it happens to my previous simulations is that,
> when I ran a trial of 20 ns, the protein moves a lot (in a cubic box with
> 1.0 nm of cutoff), and, at the final, it goes partially outside the box. I
> know that it has to move a little, but all of this movement is normal?
> Guillem
> ________________________________________
> De: Dries Van Rompaey [dries.vanrompaey at gmail.com]
> Enviat el: dimecres, 24 / febrer / 2016 7:41
> Per a: Guillem Prats Ejarque
> Tema: Re:Re: Protein-ligand system simulations with GROMACS
> Hi Guillem,
> It's best to keep everything on the mailing list, so that other people
> with your problem can find all of the posts by googling.
> You could try centering with editconf explicitly, with the -c flag. Your
> distance (-d) to the box should be bigger, though. Your cutoffs will likely
> be around 1 nm, so I'd recommend a -d of at least that much, in order to
> avoid violation of the periodic image convention.
> If you haven't already, I would take a look at Justin Lemkul's Lysozyme
> tutorial. He goes into great detail on how to set up and solvate your box.
> As for NVE/NPT/NVT, that again depends on what you want to observe. NPT is
> probably the most commonly used setup if you just want to see dynamics and
> such.
> Good luck
> Dries
> On 24 Feb 2016 12:54 a.m., "Guillem Prats Ejarque" <
> Guillem.Prats.Ejarque at uab.cat<mailto:Guillem.Prats.Ejarque at uab.cat>>
> wrote:
> Dear Dries,
> Thanks a lot for your quick response. In the literature I normally saw
> runs about one hundred nanoseconds, as you said, but I also saw some runs
> from fs to μs, so I was a little confused. I will check the paper about the
> field force, as you recommend.
> Regarding the dodecahedron box, first I tried to use it, but the protein
> does not center in the box. I do not think that this is very important, due
> to the PBC, but I found it strange, because if I use a cubic box, the
> protein is very well centered. I prepared the box like this:
> editconf -bt dodecahedron -d 0.2 -f complex2.pdb -o complex_solv.pdb
> genbox -cp complex_solv.pdb -cs spc216.gro -o complex_ion.pdb -p
> complex.top
> Looking the literature, I have doubts about NVE/NVT/NPT conditions. In
> some tutorials that I checked, with similar systems (protein-ligand
> systems) I found that they do the production run in NPT conditions, and the
> equilibration with NVT/NPT. However, in some papers, they do the whole run
> and equilibration in NVE conditions. Are there any rules to decide it, at
> least in protein-ligand systems?
> Thanks a lot, and sorry for the inconveniences,
> PS: Should I reply to the mail list instead of replying to you directly? I
> have never used a mail list before.
> Guillem
> --
> Gromacs Users mailing list
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.

More information about the gromacs.org_gmx-users mailing list