[gmx-users] solvent evaporation modeling

Tue Dec 3 23:12:42 CET 2019

Hi Sako

I did something similar where I dehydrated a system of water molecules, in
my case a crystal but the principles are the same.

I have uploaded the script at
https://github.com/aslarsen/gromacs_dehydration/blob/master/gromacs_dehydration_MPI_V2.py

It was used in this paper
https://pubs.acs.org/doi/abs/10.1021/acs.cgd.7b00889

I used gromacs 5.1.4 so you might have to update it to your version and set
whatever parameters you want.

Best Wishes
Anders

On Tue, Dec 3, 2019 at 10:36 PM Kalil Bernardino <kalil.bernardino at gmail.com>
wrote:

>  Dear Sako,
>
>
> Unfortunately I don't have access to the script in the present since I'm in
> the middle of a travel. But anyway it was done in a specific way for the
> system that we were working on that time and would need some changes to
> work with different systems.
>
> I can describe to you the general idea here.
>
>
> First, there is two ways to simulate the solvent evaporation and we
> performed both in the paper André cited. The more physical way, since we
> were working with a liquid vacuum interface, suppose perpendicular to z
> direction, was to run a short simulation, save the final structure, check
> which molecules are above some zcut value, exclude they from your
> structure, rewrite the topology file to update the number of molecules, run
> the grompp to produce the new tpr and run a new short simulation. This
> would be something similar to have a liquid with a vacuum pump at z = zcut,
> and any molecule that enters the vacuum pump will be removed from the
> system. For this protocol, you should work with constant volume
> simulations.
>
>
> The problem with this procedure is that it can be very slow, specially when
> you have too few solvent remaining or if your system have a small vapor
> pressure, as is your case since you are working with water. So, if you want
> to complete dry your system, you can remove molecules randomly in the
> following way. You may or may not want to create a liquid/vacuum interface,
> depending if you want to simulate the molecule in some interface or if you
> want to simulate the bulk of something that in being dried. If you have a
> vacuum interface, you need to work with NVT simulations, but in order to
> make a bulk simulation you should do in a NPT ensemble.
>
>
> 1- Equilibrate your system with the solvent
>
> 2- Save the final structure
>
> 3- Delete some fraction of randomly selected solvent molecules. You can do
> this directly with some script or can use the gromacs command genion to
> randomly convert some molecules in some monoatomic specie (for instance,
> "genion -np 60 -pname DEL" and select the water for replace 60 water
> molecules with 60 particles with name DEL), make_ndx to generate a index
> file with a group including every molecules except the specie produced by
> genion (DEL in the example) and then the editconf reading the index.ndx in
> order to produce a new .gro file with every molecule except the deleted
> ones. This .gro will be your initial structure for next step. If you want
> to use the gromacs commands for this remotion, remember to run the grompp
> again from your final structure to produce a new tpr file before using
> genion.
>
> 4- Rewrite your topology file in order to update the number of solvent
> molecules
>
> 5- Run grompp and mdrun to do a short simulation with this small numer of
> water molecules and save the final structure
>
> 6- If still have water in your system (or if have more water than you want
> in your final structure), go back to step 3 and continue removing.
>
>
> You should not remove a large amount of molecules in a single step or this
> can lead to some artificial structures. And by the end of your process,
> when you have only a small amount of solvent, I think it is better to
> reduce even more the amount removed at each step (which is not so bad
> because your simulation will run faster at this point).
>
>
> Hope this protocol helps you.
>
>
> Best,
>
> Kalil
>
>
>
> Em seg., 2 de dez. de 2019 às 14:14, SAKO MIRZAIE <sako.biochem at gmail.com>
> escreveu:
>
> > Dear André,
> >
> > Thank you very much for your email and hope Kalil has the script.
> >
> > Best regards
> >
> > On Mon, Dec 2, 2019 at 7:15 AM André Farias de Moura <moura at ufscar.br>
> > wrote:
> >
> >> Dear Sako,
> >>
> >> I'm ccing Kalil, who actually wrote and run the script to remove solvent
> >> molecules, he might still have it.
> >>
> >> regards
> >>
> >> Andre
> >>
> >> On Mon, Dec 2, 2019 at 2:54 AM SAKO MIRZAIE <sako.biochem at gmail.com>
> >> wrote:
> >>
> >>> Dear André,
> >>>
> >>> Thank you for your response.
> >>> Could you send me such a script?
> >>>
> >>> On Sun, Dec 1, 2019 at 7:15 PM André Farias de Moura <moura at ufscar.br>
> >>> wrote:
> >>>
> >>>> Dear Sako,
> >>>>
> >>>> we did something like that a few years ago, please take a look at DOI
> 10.1039/C4CP03519D
> >>>> for details.
> >>>>
> >>>> in a nutshell: you need a script that runs a sequence of short
> >>>> equilibration and production runs after a number of solvent molecules
> are
> >>>> removed (implying that topology needs to be updated for the number of
> >>>> solvent molecules at each round, so the script needs to include some
> >>>> parsing of the files as well).
> >>>>
> >>>> Andre
> >>>>
> >>>> On Sun, Dec 1, 2019 at 8:03 PM SAKO MIRZAIE <sako.biochem at gmail.com>
> >>>> wrote:
> >>>>
> >>>>> Hi All,
> >>>>>
> >>>>> I want to simulate a polymer: protein system in a way that water
> >>>>> solvent
> >>>>> will evaporated gradually. How should I do that? What parameters are
> >>>>> needed
> >>>>> to be included in the mdp file.
> >>>>>
> >>>>> Best
> >>>>>
> >>>>>
> >>>>>
> >>>>> --
> >>>>> ***********************************************
> >>>>> Sako
> >>>>> --
> >>>>> 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.
> >>>>>
> >>>>
> >>>>
> >>>> --
> >>>> _____________
> >>>>
> >>>> Prof. Dr. André Farias de Moura
> >>>> Department of Chemistry
> >>>> Federal University of São Carlos
> >>>> São Carlos - Brazil
> >>>> phone: +55-16-3351-8090
> >>>>
> >>>
> >>>
> >>> --
> >>> ***********************************************
> >>> Sako Mirzaie
> >>> Sako Mirzaie
> >>> Ph.D. in biochemistry, Assistant Professor, Science Faculty, Islamic
> >>> Azad University of Sanandaj, Sanandaj, Iran
> >>>
> >>> Visiting Professor, Advanced Pharmaceutics
> >>>
> >>> & Drug Delivery Laboratory
> >>>
> >>> Leslie Dan Faculty of Pharmacy
> >>>
> >>> University of Toronto
> >>>
> >>> 144 College Street, Toronto, Ontario
> >>>
> >>> Canada M5S 3M2
> >>>
> >>> http://scholar.google.com/citations?user=viwZvVAAAAAJ&hl=en
> >>>
> >>> http://www.scopus.com/authid/detail.url?authorId=54886431500
> >>>
> >>> http://www.ncbi.nlm.nih.gov/pubmed/?term=sako+mirzaie
> >>> https://www.researchgate.net/profile/Sako_Mirzaie/publications/
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>
> >> --
> >> _____________
> >>
> >> Prof. Dr. André Farias de Moura
> >> Department of Chemistry
> >> Federal University of São Carlos
> >> São Carlos - Brazil
> >> phone: +55-16-3351-8090
> >>
> >
> >
> > --
> > ***********************************************
> > Sako Mirzaie
> >
> >
> >
> >
> >
> >
> >
> >
> --
> 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.