[gmx-users] crash due to huge files using reaction field

[hverli at cbiot.ufrgs.br]

Hi,

I have been experiencing some strange problems with reaction field. In two
different protein systems with about ~120 aminoacids each one (including water
and ions), at some point of the simulations (~4.0ns) gromacs generated a huge
.log file (>60GB!), completely filling the hard disk and so stopping the
simulation. We tried to simulate those systems in other machines, considering a
hardware problem, but the error still occurs (approximately in the same time
step). We are using the Mandrake linux, and so considering a possible problem
with the operational system we simulated in both Makdrake 9.2 and 10.0, and
again the error was observed. It appears to be quite reproducible.

In other smaller system (~30 aminoacids), the simulation proceeds fine (until
10.0ns), using the same .mdp file (reproduced in the text).

tinit               =  4000.0
dt                  =  0.002	; ps !
nsteps              =  200000	; total 4000-4400 ps.
nstcomm             =  1
nstxout             =  250
nstvout             =  1000
nstfout             =  0
nstlog              =  100
nstenergy           =  100
nstlist             =  10
ns_type             =  grid
coulombtype         =  Generalized-Reaction-Field
rlist               =  1.0
rcoulomb            =  1.6
rvdw                =  1.6
optimize_fft	    =  yes
epsilon_r           =  54.0
; Berendsen temperature coupling is on in four groups
Tcoupl              =  berendsen
tc-grps		    =  Protein SOL Cl
tau_t               =  0.1 0.1 0.1
ref_t               =  310 310 310
; Energy monitoring
energygrps          =  Protein SOL Cl
; Isotropic pressure coupling is now on
Pcoupl              =  berendsen
Pcoupltype          = isotropic
tau_p               =  0.5
compressibility     =  4.5e-5
ref_p               =  1.0
; Generate velocites is off at 100 K.
gen_vel             =  no
gen_temp            =  310.0
gen_seed            =  173529

Is there some cause for such strange error? Should I try to increase the cutoffs
or change other parameters in the mdp file?

Many thanks in advance,

Hugo Verli.


--------------------------------------------------------------------
Centro de Biotecnologia da Universidade Federal do Rio Grande do Sul