[gmx-users] Re: Energy minimization problem(1-4 interaction at distance 3.540 is larger than the 1-4 table size 1.000 nm)
arun.target at gmail.com
Sat May 5 20:09:38 CEST 2007
Hi Tsjerk and others,
I tried to simulate one surfactant and one
chloride ion in 500 waters and got similar type of errors. I mean my
system got collapsed for l-bfgs minimization and for steepest descent
the potential energy is a large number.(obviously some bad contacts).
It seems the problem is with topology. Earlier I asked about what is
meant by broken topology?? Can anyone tell what may ne the problem and
also about what is meant by broken topology??
Thanks and Regards
M.E chem engg
> Hi Tsjerk and gmx-users,
> Thank you for the response. And I will be more
> clear in asking questions. The force field I am using is ffgmx. I am
> not sure I will obtain good properties for my system with ffgmx at
> this time. But I want to understand how to set up my simulation which
> contains ionic surfactants.
> And yes I will strat it for one surfactant and
> one counterion in water.
> In my post, MDRUN means True dynamics( I mean
> real md not energy minimization). I know that without minimizing the
> energy it will be bad to do MD. But I just checked how it will work as
> I kept density around 330 kg/m^3. I think it won't work even. Sorry
> for confusing with last post.
> My system's initial configuration is alright it
> seems(No overlapping atoms and bonds). And after I start EM or MD it's
> exploding(Why I don't know exactly).
> Now I will tell the steps I took to start my
> simulation. Initially I got topology and pdb files(united atom) from
> PRODRG(gromos 87 force field) for my surfactant behenyl(C22) trimethyl
> ammonium chloride(charge +1, it won't contain chloride) and stearyl
> alcohol. And then I got water.pdb,chlorine.pdb from pdb database. I
> prepared a random configuration of 1400 water molecules, 20 surf, 20
> chloride ions and 40 co-surfs in 5X5X8 nm^3 by using packmol.
> Then I used editconf to get system.gro from
> system.pdb and also put in a cubic box with this command
> editconf -f system.pdb -bt cubic -o system.gro
> But it reported system volume was zero. So I changed
> end of my .gro file by replacing 0.0000 0.0000 0.0000 by 5.000 5.000
> 8.000. The density is about 330 kg/m^3. So I thought it will be useful
> in Energy Minimization and also can be adjusted in NPT md simulation.
> As i have a little experience in DPPC topologies and running
> simulation for DPPC bilayer phase(successfully) using the files got
> from Tieleman's website, I prepared a topology file for my
> system(btm.top). It contains the following lines
> #include "ffgmx.itp"
> #include "btm87.itp"
> #include "stearyl87.itp"
> #include "ions.itp"
> #ifdef FLEX_SPC
> #include "flexspc.itp"
> #include "spc.itp"
> [ system ]
> BTM with cosurf
> [ molecules ]
> ; name number
> BTM 20
> STR 40
> SOL 1400
> Cl 20
> btm87.itp(surf) and stearyl87.itp(co-surf) are topologies got from PRODRG.
> I kept all these in my direactory. Now I tried to minmize
> the energy of the system. First I used grompp to create md input
> file(I mean for EM). It worked fine. But for mdrun system is
> exploding(for l-bfgs method). For steepest descent it's not, But the
> potential energy of the system was a big positive number at end. For
> my true MD also the system's exploding like l-bfgs minmization after
> showing that warning earlier I posted. These are the steps I took and
> results I got. And for last query regarding tc-grps, I coupled the
> whole system to a heat bath at 300 K. I observed the previous posts.
> Here with this mail I attached my MD .mdp file.
> And at last Is there any relation with Mark's(Mark abraham)
> observations regarding broken topologies in some of the posts I
> observed, with my problem. Can anyone explain what is meant by a
> broken topology file. Try to help me to understand these thing
> I can give any more details if it is needed.
> Thanks and regards
> Arun Kumar
> M.E chem engg
> > Hi Arun,
> > You have to get your methods straight.., and be more clear in the
> > phrasing of your question.
> > > I am trying to simulate a surfactant-water
> > > system(also contains some co-surfactants). Surfactant is a cationic
> > > surfactant with chloride as counterion. And the cosurfactant is fatty
> > > alcohol. water is of spc model. I got topologies from Prodrg for my
> > > surfactant and cosurfactant.
> > What force field do you use? The standard output from PRODRG for
> > gromacs is the gmx force field, which is deprecated and shouldn't be
> > used (this is in the archives).
> > > So I minimized the energies for both my
> > > molecules. And prepared a system of 1400 waters, 20 surfactants, 20 counter
> > > ions and 40 cosurfactants in a box of 5X5X8 nm^3 using packmol.
> > >
> > Maybe you're better of trying one surfactant molecule and one chloride
> > in water to start with?
> > > I tried to minimize the energy of the system using
> > > l-bfgs. But I am always getting the error attached below. In my em.mdp file
> > > I am using pme for coulomb type. Even for MDRUN also I am getting the same
> > > error... I checked the previous posts and I am unable to understand what's
> > > the problem. And can anyone tell what is meant by broken topology??? Also
> > > try to help me to get out of the problem.
> > Even for MDRUN? pfff, that must be really bad then. First of all, do
> > you mean running the program 'mdrun' with the .tpr file for l-bfgs EM,
> > or an actual MD run? In case of the latter, wouldn't you think there's
> > little point in starting an MD run if energy minimization already
> > failed? In case of the former, the statement is redundant (now it's
> > just confusing).
> > > F-max = inf on atom 97
> > > F-Norm = nan
> > > -----------
> > So, your system is exploding... Do you have any overlapping
> > bonds/atoms? Does your box actually contain your system? (Do think
> > periodically..., overlapping periodic images?).
> > >
> > > My topology contains a system of four types surfactant, cosurfctant, water
> > > and Chloride ions. Until grompp everything is fine. Total charge on system
> > > is zero. I can give any more details if it is needed.
> > Good, until grompp everything's fine. Now, then the error comes as a
> > surprise! But what did you do before? What steps did you take?
> > Oh, and I bet that in your .mdp file it says "tc-grps = SOL SUR ALC
> > CL", albeit that the actual names and the order of them may be
> > different. This is an issue, which has also been dealt with
> > extensively (including the past week). In any case, when reporting
> > that you've ran into a problem, it may be helpful to us to send your
> > .mdp file.
> > Cheers,
> > Tsjerk
> > --
> > Tsjerk A. Wassenaar, Ph.D.
> > Junior UD (post-doc)
> > Biomolecular NMR, Bijvoet Center
> > Utrecht University
> > Padualaan 8
> > 3584 CH Utrecht
> > The Netherlands
> > P: +31-30-2539931
> > F: +31-30-2537623
> -------------- next part --------------
> A non-text attachment was scrubbed...
> Name: btm.mdp
> Type: application/octet-stream
> Size: 5284 bytes
> Desc: not available
> Url : http://www.gromacs.org/pipermail/gmx-users/attachments/20070505/46e7b66d/btm-0001.obj
More information about the gromacs.org_gmx-users