From: Rajan Vatassery (rajan_at_umn.edu)
Date: Mon Sep 03 2012 - 15:03:21 CDT
Amin,
If you have minimized for more than 500 steps and your vdW energies are
still at 99999999, then it's possible you have superimposed atoms on top
of each other. If there is no difference in the coordinates of atoms,
then there is no gradient and therefore the minimizer will not have any
effect. I've mistakenly created simulations in which two solvent
molecules are superimposed and this leads to the same errors you speak
of.
The log file should tell you which atom is moving to fast (or which
atom has a SHAKE error), so check to see if another atom has the same
coordinates in the PDB file.
On a related note, I think VMD could add an error message in cases
where atoms are superimposed. While it wouldn't be widely useful, it
would be easy enough to implement that it would be worthwhile
expenditure of time.
rajan
On Mon, 2012-09-03 at 22:20 +0530, amin_at_imtech.res.in wrote:
> Thanks. I checked with dynamic bonds. I find that when I wrap using the
> dimensions of water, the layer of water below the membrane disappears and the
> protein becomes fragmented into four parts but the water box appears as one.
> However, when I wrap using the dimensions of the whole system, which includes
> POPC membrane, the box gets split into 8 pieces with cuts parallel to z and x
> axis. Also I found that due to my mistake, my box dimensions in the namd config
> file were smaller. I increased them according to the size of the whole system,
> and now the simulation runs into the first equilibration stage. However, the vdw
> energies are still 99999999 even when the equilibration run is going on.
>
> Regards.
> Amin.
>
>
>
> > Hi,
> >
> > the long bonds are normal, but you should check with dynmic bonds and low cutoff
> > then to see if the box you set was to small. I can't follow your script now
> > (time), but why don't you use the box you set when solvating??
> >
> > For better view, but maybe not reliable is to do the wrap with -compound res
> >
> > Norman Geist.
> >
> >> -----Ursprüngliche Nachricht-----
> >> Von: owner-namd-l_at_ks.uiuc.edu [mailto:owner-namd-l_at_ks.uiuc.edu] Im
> >> Auftrag von amin_at_imtech.res.in
> >> Gesendet: Montag, 3. September 2012 13:27
> >> An: Norman Geist
> >> Cc: ajasja.ljubetic_at_gmail.com; namd-l_at_ks.uiuc.edu
> >> Betreff: Re: AW: AW: namd-l: Error in membrane protein simulation
> >>
> >> Thanks. I tried as you said and after pbc wrap all I just see a lot of
> >> long
> >> lines crossing each other and everything is cluttered together. I used
> >> the
> >> following script to calculate the cell dimensions
> >>
> >> set sel [atomselect top all]
> >> echo [measure center $sel]
> >> set m [measure minmax $sel] ;foreach {j1 j2} $m {} ;foreach {x2 y2 z2}
> >> $j2 {}
> >> ;foreach {x1 y1 z1} $j1 {}
> >> set x [expr $x2 - $x1]
> >> set y [expr $y2 - $y1]
> >> set z [expr $z2 - $z1]
> >> echo $x $y $z
> >>
> >> Can you please tell whats wrong here?
> >>
> >> Warm regards.
> >> Amin.
> >>
> >> > Ok, another thing you should try is:
> >> >
> >> >
> >> >
> >> > 1. Load the systems psf and pdb
> >> >
> >> > 2. Set the box dimensions you used in namd with pbc set {x y z}
> >> >
> >> > 3. Enter pbc wrap
> >> >
> >> > 4. Check if there are overlapping atoms now
> >> >
> >> >
> >> >
> >> > This will show if your box dimensions in XY are correct, if not, you
> >> will see a
> >> > lot close atoms.
> >> >
> >> >
> >> >
> >> > Mit freundlichen Grüßen
> >> >
> >> >
> >> >
> >> > Norman Geist.
> >> >
> >> >
> >> >
> >> > Von: owner-namd-l_at_ks.uiuc.edu [mailto:owner-namd-l_at_ks.uiuc.edu] Im
> >> Auftrag von
> >> > Ajasja Ljubetic
> >> > Gesendet: Montag, 3. September 2012 11:16
> >> > An: amin_at_imtech.res.in
> >> > Cc: Norman Geist; namd-l_at_ks.uiuc.edu
> >> > Betreff: Re: AW: namd-l: Error in membrane protein simulation
> >> >
> >> >
> >> >
> >> > One thing you can try is setting a very large langevin damping, so
> >> that the
> >> > atoms don't explode and them minimize and run and minimize and run.
> >> Also atoms
> >> > with unknown positions get set to {0 0 0} so check if you have
> >> multiple atoms at
> >> > the origin.
> >> >
> >> >
> >> >
> >> > Regards,
> >> >
> >> > Ajasja
> >> >
> >> > Example setting:
> >> >
> >> > langevinDamping 100;# or even 1000
> >> >
> >> > langevinHydrogen yes
> >> >
> >> >
> >> >
> >> > minimize 500
> >> >
> >> > run 50
> >> >
> >> > minimize 250
> >> >
> >> > run 50
> >> >
> >> > minimize 100
> >> >
> >> > run 100
> >> >
> >> > #etc ...
> >> >
> >> >
> >> >
> >> > On 3 September 2012 10:41, <amin_at_imtech.res.in> wrote:
> >> >
> >> > Thanks for the reply. Yes 9999999 is VDW. I checked for the
> >> overlapping atoms
> >> > and there are no ions within the region of the membrane. However the
> >> water
> >> > molecules come very slightly below the P1 atoms of the membrane but
> >> it seems
> >> > that they are coming from the spaces between the lipid molecules. Can
> >> these
> >> > water atoms be causing the trouble? Can you please tell me what else
> >> I can look
> >> > for?
> >> >
> >> > Regards.
> >> > Amin.
> >> >
> >> >
> >> >
> >> >> Hi,
> >> >>
> >> >> as you did unfortunately not mention which of the energies are
> >> 9999999, I can
> >> >> only guess.
> >> >>
> >> >> So I guess it is the VDW energy. If I guess correct, you have
> >> overlapping
> >> >> atoms.
> >> >> The minimizer will mostly never get rid of such bad initial
> >> conformations. So
> >> >> you need to find the bad contacts (maybe using vmd and dynamic bonds
> >> with a
> >> >> low
> >> >> cutoff), or just rebuild the system. As you told you have a
> >> membrane, could it
> >> >> be that you have ions beside the membrane? Remember periodicity and
> >> that these
> >> >> ions are within the membrane then, as they get wrapped at simulation
> >> start.
> >> >>
> >> >> regards
> >> >>
> >> >> Norman Geist.
> >> >>
> >> >>
> >> >>> -----Ursprüngliche Nachricht-----
> >> >>> Von: owner-namd-l_at_ks.uiuc.edu [mailto:owner-namd-l_at_ks.uiuc.edu] Im
> >> >>> Auftrag von amin_at_imtech.res.in
> >> >>> Gesendet: Samstag, 1. September 2012 08:59
> >> >>> An: namd-l_at_ks.uiuc.edu
> >> >>> Betreff: namd-l: Error in membrane protein simulation
> >> >>>
> >> >
> >> >>> Dear all,
> >> >>> I am trying to simulate a protein embedded in a POPC
> >> membrane. I
> >> >>> am
> >> >>> following the NAMD tutorial on membrane proteins. When I go to the
> >> >>> first
> >> >>> equilibration step using constraints mentioned in the step "melting
> >> of
> >> >>> membrane" I get "Constraints failure in rattle algorithm" with
> >> >>> rigidbonds
> >> >>> all and "Atoms moving too fast" with rigidbonds water. I looked for
> >> the
> >> >>> solutions in mailing list and tried 1) decreasing time step to 0.5,
> >> 2)
> >> >>> increasing minimization steps from 1000 to 10000, 3) increasing
> >> >>> pairlistdist to 16, 4) setting margin upto 10, 5) gradually heating
> >> at
> >> >>> a
> >> >>> slow rate. Also I noticed that during minimization I have energies
> >> >>> 9999999. I have checked the atoms mentioned in the error file
> >> visually
> >> >>> and they are distributed throughout the membrane.There are more
> >> than
> >> >>> 5000
> >> >>> of them. I made the membrane using VMD membrane plugin. Can someone
> >> >>> please tell me where I might be making an error.
> >> >>>
> >> >>> Regards.
> >> >>>
> >> >>> Amin.
> >> >>>
> >> >>>
> >> ______________________________________________________________________
> >> >>> सूक्ष्मजीव प्रौद्योगिकी संस्थान (वैज्ञानिक
> >> औद्योगिक
> >> >>> अनुसंधान परिषद)
> >> >>> Institute of Microbial Technology (A CONSTITUENT ESTABLISHMENT OF
> >> CSIR)
> >> >>> सैक्टर 39 ए, चण्डीगढ़ / Sector 39-A, Chandigarh
> >> >>> पिन कोड/PIN CODE :160036
> >> >>> दूरभाष/EPABX :0172 6665 201-202
> >> >>
> >> >>
> >> >>
> >> >
> >> >
> >> >
> >> ______________________________________________________________________
> >> > सूक्ष्मजीव प्रौद्योगिकी संस्थान (वैज्ञानिक औद्योगिक
> >> अनुसंधान परिषद)
> >> > Institute of Microbial Technology (A CONSTITUENT ESTABLISHMENT OF
> >> CSIR)
> >> > सैक्टर 39 ए, चण्डीगढ़ / Sector 39-A, Chandigarh
> >> > पिन कोड/PIN CODE :160036
> >> > दूरभाष/EPABX :0172 6665 201-202
> >> >
> >> >
> >> >
> >> >
> >>
> >> ______________________________________________________________________
> >> सूक्ष्मजीव प्रौद्योगिकी संस्थान (वैज्ञानिक औद्योगिक
> >> अनुसंधान परिषद)
> >> Institute of Microbial Technology (A CONSTITUENT ESTABLISHMENT OF CSIR)
> >> सैक्टर 39 ए, चण्डीगढ़ / Sector 39-A, Chandigarh
> >> पिन कोड/PIN CODE :160036
> >> दूरभाष/EPABX :0172 6665 201-202
> >
> >
> >
>
> ______________________________________________________________________
> सूक्ष्मजीव प्रौद्योगिकी संस्थान (वैज्ञानिक औद्योगिक अनुसंधान परिषद)
> Institute of Microbial Technology (A CONSTITUENT ESTABLISHMENT OF CSIR)
> सैक्टर 39 ए, चण्डीगढ़ / Sector 39-A, Chandigarh
> पिन कोड/PIN CODE :160036
> दूरभाष/EPABX :0172 6665 201-202
>
This archive was generated by hypermail 2.1.6 : Tue Dec 31 2013 - 23:22:30 CST