[gmx-users] LJ-14 energy

Ming Tang m21.tang at qut.edu.au
Sat Jul 18 12:26:34 CEST 2015 

Dear Mark,

Thanks for your link. Now I have a better understanding on cutoff-scheme = verlet. This also let me notice that there are difference between version 5.0.4 and the earlier versions. Actually I came across another simulation problem, which made me feel confused for a long time. I saw several publications, in which people restrained the COM of one group (referred as reference group) in SMD with version 4.6, but I did not find the way to achieve this. both freezing group and restraining all the atoms in the group fix the atoms not COM. Do you have any suggestion for me? 

Thanks.

-----Original Message-----
From: gromacs.org_gmx-users-bounces at maillist.sys.kth.se [mailto:gromacs.org_gmx-users-bounces at maillist.sys.kth.se] On Behalf Of Mark Abraham
Sent: Friday, 17 July 2015 5:21 AM
To: gmx-users at gromacs.org
Subject: Re: [gmx-users] LJ-14 energy

On Thu, Jul 16, 2015 at 1:24 PM Ming Tang <m21.tang at qut.edu.au> wrote:

> Dear Justin,
>
> Thanks.
> 1.  The total Lennar-Jones 12-6 energy of the whole system is the 
> addition of LJ-14, LJ-SR and LJ-LR. Analogously, the coulomb energy is 
> the addition of coulomb-14, coulomb-SR and coulomb-LR. Is my 
> understanding of your explanation right? Besides, in Gromacs, is the 
> potential energy the addition of G96Bond, G96Angle, Proper-Dih, 
> Improper-Dih, LJ-14, Coulomb-14, LJ-SR, LJ-LR, Coulomb-SR and Coulomb-LR?
> You are right. That's why I don't get the LJ-LR and Coulomb-LR options 
> when using g_energy. Suddenly, I realised that my understanding of the 
> twin-range approach was totally wrong. For cutoff-scheme = verlet, 
> rlist cannot be smaller than rvdw = rcoulomb, and When using 
> twin-range method, either coulombtype = cut-off with rcoulomb ≥ rlist 
> or vdwtype = cut-off with rvdw ≥ rlist. Does this mean that for 
> twin-range approach, rlist = rcoulomb = rvdw when cutoff-scheme = 
> verlet? However, even if I set vdw-modifier = potential- switch, the 
> rlistlong larger than rlist is reset to be equal to rlist by gromacs, 
> and there is not LJ-SR option when using g_energy. I am totally lost. 
> Can you help to tell me how does twin-range work when cutoff-scheme = verlet?
>

http://manual.gromacs.org/documentation/5.1-rc1/user-guide/cutoff-schemes.html
(and equivalent docs for earlier versions) indicate that twin-range is not supported by the Verlet scheme.

(Technically, it could be supported, and some forms of it are in fact used by the PME-tuning facility, but some of us are unconvinced of the technical merit of such forms, nor do we like the confusion created by how such group-scheme features are named and implemented in GROMACS. So, if there is ever support, it's likely to be from a clean slate, and after we remove the group scheme.)

Another question is I used the following code to make my cut-off parameters
> be consistent with those used for the triple-range cut-off scheme in 
> the paper defining Gromos 54a7 force field:
>
> coulombtype      =  reaction-field
> coulomb-modifier =  potential-shift
> rcoulomb-switch  =  0.8
> rcoulomb         =  1.4
> epsilon_rf       =  61
>
> Currently, I think it is wrong. It seems that rlist = 0.8 is reasonable.
> But rlist should not be smaller than rcoulomb when cutoff-scheme = verlet.
> How to achieve this triple-range cut-off scheme?
>

You can't.

Mark


> 2.   The pull rate I use is 4e-5/ps. I thought it is small so that the
> energy of water can be treated as constant during the pull process. I 
> forgot to consider the water-protein interaction and that my pull rate 
> is much larger than the pull rates used in the experiments which is in 
> order of micrometre per second and is much more relevant to 
> physiological pulling rate. You are right, I cannot treat the trends 
> of the system's energy as those of the protein's energy. But how can I get the energy of the protein?
> Is it feasible to set all the atoms of the protein as one energy 
> group, and do rerun to achieve this?
>
> Thanks,
> Regards
> -----Original Message-----
> From: gromacs.org_gmx-users-bounces at maillist.sys.kth.se [mailto:
> gromacs.org_gmx-users-bounces at maillist.sys.kth.se] On Behalf Of Justin 
> Lemkul
> Sent: Thursday, 16 July 2015 11:20 AM
> To: gmx-users at gromacs.org
> Subject: Re: [gmx-users] LJ-14 energy
>
>
>
> On 7/15/15 8:39 PM, Ming Tang wrote:
> > Dear Chaban and Justin,
> >
> >
> >
> > Sorry for the mis-action. Please ignore my last email.
> >
> >
> >
> > Thank you both for your free help.
> >
> > I am pulling a collagen triple helix in a periodic water box using 
> > umbrella direction-periodic. I want to calculate the Lennard-Jones
> > 12-6 of the protein. According to Gromacs manual, it contains 
> > repulsive short-range term and attractive long-range term. Is the 
> > Lennard-Jones 12-6 energy simply the addition of LJ-SR and LJ-LR? I 
> > am interested in the energy trends more than
>
> LJ-LR only exists with a twin-range cutoff approach.  The total LJ 
> between any groups also includes LJ-14 for intramolecular terms.  
> Intermolecular LJ would be the sum of LJ-SR and LJ-LR, if the latter exists.
>
> > the energy absolute values. So, if I can calculate the LJ 12-6 of 
> > the whole system, it is fine, because I can assume the energy of the 
> > water is not subject to change due to pulling.
> >
>
> Why can you assume that?  You're pulling a structure, which is a 
> non-equilibrium process.  There will be different interactions as a 
> function of time.
> Water-water interactions at long distance from the protein indeed are 
> unlikely to be affected, but those close to the protein may be 
> different (though electrostatics/hydrogen bonding are probably much more interesting).
> Water-protein interactions certainly can't be assumed to be unchanging 
> during pulling.
>
> -Justin
>
> --
> ==================================================
>
> Justin A. Lemkul, Ph.D.
> Ruth L. Kirschstein NRSA Postdoctoral Fellow
>
> Department of Pharmaceutical Sciences
> School of Pharmacy
> Health Sciences Facility II, Room 629
> University of Maryland, Baltimore
> 20 Penn St.
> Baltimore, MD 21201
>
> jalemkul at outerbanks.umaryland.edu | (410) 706-7441 
> http://mackerell.umaryland.edu/~jalemkul
>
> ==================================================
> --
> 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.
> --
> 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.
--
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