[gmx-users] Error during mdrun

[Mark Abraham]

chetana baliga wrote:
> Thanks, Tsjerk and Mark.  Here's what the end of my log file says :
> 
>           Step           Time         Lambda
>        5243900    10487.80078        0.00000
> 
>   Rel. Constraint Deviation:  Max    between atoms     RMS
>       Before LINCS         0.019872     61     62   0.004609
>        After LINCS         0.000051     35     37   0.000020
> 
>   Energies (kJ/mol)
>       G96Angle    Proper Dih.  Improper Dih.          LJ-14     Coulomb-14
>    1.58108e+02    9.30052e+01    4.86079e+01    6.91807e+00    1.08849e+03
>        LJ (SR)   Coulomb (SR)   Coulomb (LR)      Potential    Kinetic En.
>    3.04429e+04   -1.88462e+05   -1.73048e+04   -1.73929e+05    3.14049e+04
>   Total Energy    Temperature Pressure (bar)
>   -1.42524e+05    3.00225e+02    3.27136e+01
> 
> 
> t = 10487.916 ps: Water molecule starting at atom 6211 can not be settled.
> Check for bad contacts and/or reduce the timestep.Wrote pdb files with
> previous and current coordinates
> Large VCM(group rest):   -440.66449,      0.00000,   -110.16612,
> ekin-cm:  7.85519e+09
> 
> My peptide is a decapeptide, and yes, it does contain a lysine, but
> the error doesnt involve its atoms.  I am running my simualtions under
> npt conditions. I initially did a pr run of a total of 600 ps, initial
> 200ps being under nvt conditions and the remaining under npt [as
> suggested in the archives, that initial nvt run helps the system to
> relax] After the pr, the system had equilibrated quite well, though
> the pressure average was 5.85 bar , while the ref_p was 1 bar. The
> solvent used is SPC water, and the coulomb type is PME.
> 
> And, there were no other LINCS warnings in the md0.log file.
> 
> Could you please advise me on this?

OK, the water "can not be settled" is a reference to SETTLE, which is a 
constraint-enforcing algorithm for water... so we were on the right 
track with the suspicion of LINCS errors (another such algorithm).

This time step is quite a long time away from your equilibration time 
(assuming you set t=0 at the start of equilibration) for such an error 
to arise. I would first have a look at your trajectory just before that 
time and see if you can deduce anything from the behaviour, especially 
of and near atom 6211.

Over what length of time did you observe the average pressure of 5.85 
bar? You shouldn't be including the initial period, since it almost 
certainly was very large then, and will be throwing off your statistics. 
If you have xmgrace installed, use g_energy -W to look at a graphical 
plot of pressure with time in the both the NPT equilibration and the NPT 
production run.

Mark