[gmx-users] Water molecule starting at atom X can not be settled; can't find a solution

[Simone Cirri]

Hi everyone,

in the last weeks I've been trying to run a simulation with Gromacs 4.0.5
and the force field ffamber99.
The protein is albumine (BSA, bovine serum albumine), so it's a very big one
(570+ residues).
The solvent is tip3p.
After an energy minimization run and a position restraint run, I start with
the "real" simulation. Invariably, it crashes after 1 ps with the error
message:

t = 1.000 ps: Water molecule starting at atom 42031 can not be settled.
Check for bad contacts and/or reduce the timestep.

I've searched in the mailing list and I've tried many of the solutions you
proposed:

- I looked at the water molecule with the problematic atom to see if it was
interacting in some weird way with the protein (I put ntsxout = 1 in order
to have many more frames); but it isn't so, since that molecule is only one
of the many solvent molecules that (at least in the first ps) stay far from
the protein
- I tried to run a longer and stronger minimization: so I increased the
number of steps (from 10000 to 20000) and I decreased emtol (from 100 to 1),
but nothing changed
- I tried to increase the box dimensions; in editconf I put -c 1.2 (instead
of 0.7), but nothing changed.
- I checked the stepXXXb.pdb and stepXXXc.pdb that mdrun created after the
crash, but I didn't see anything strange in them.
- I did a longer position restraint run (3 ns instead of the usual 200 ps)
but nothing changed... I was hoping that it would stabilize the water, but I
think the problem is not that.

Even though the energy minimization doesn't show any problems, I think
that's the crucial part; as someone said in another message, the problem
doesn't really involve the water molecule that cannot be settled, but other
atoms in the protein with big forces that in some way collide with the water
molecule... am I right?
In this case, deleting the problematic water molecule could help? I thought
it would be useless, since that molecule is not close to the protein at
all...

Here you have the em.mdp and the md.mdp, if you want I can also paste the
pr.mdp .

title               =  hsacyx
cpp                 =  /lib/cpp
define              =  -DFLEX_SPC
constraints         =  none
integrator          =  steep
dt                  =  0.002    ; ps !
nsteps              =  20000
nstlist             =  10
ns_type             =  grid
rlist               =  0.9
coulombtype        =  PME
rcoulomb            =  0.9
rvdw                =  0.9
fourierspacing        =  0.12
fourier_nx        =  0
fourier_ny        =  0
fourier_nz        =  0
pme_order        =  6
ewald_rtol        =  1e-5
optimize_fft        =  yes
;
;       Energy minimizing stuff
;
emtol               =  1
emstep              =  0.01



title               =  hsacyx
cpp                 =  /lib/cpp
constraints         =  all-bonds
integrator          =  md
dt                  =  0.002    ; ps !
nsteps              =  7500000; total 15000 ps.
nstcomm             =  1
nstxout             =  2500
nstvout             =  0
nstfout             =  0
nstlist             =  500
ns_type             =  grid
rlist               =  0.9
coulombtype        =  PME
rcoulomb            =  0.9
rvdw                =  0.9
fourierspacing        =  0.12
fourier_nx        =  0
fourier_ny        =  0
fourier_nz        =  0
pme_order        =  6
ewald_rtol        =  1e-5
optimize_fft        =  yes
; Berendsen temperature coupling is on in three groups
Tcoupl              =  berendsen
tau_t               =  0.1      0.1      0.1
tc-grps            =  protein      sol     NA+
ref_t               =  300      300     300
; Pressure coupling is on
Pcoupl              =  berendsen
pcoupltype          =  isotropic
tau_p               =  0.5
compressibility     =  4.5e-5
ref_p               =  1.0
; Generate velocites is on at 300 K.
gen_vel             =  yes
gen_temp            =  300.0
gen_seed            =  173529


I really don't know what else I could do... if someone has an idea, I would
be glad to hear it.
Thank you

Simone Cirri
PhD student at the Structural Biology Lab of University of Siena, Italy
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20091016/95a8a86c/attachment.html>