"FATAL ERROR: Bad global improper count" revisited

From: Markus K. Dahlgren (markus.dahlgren_at_yale.edu)
Date: Tue Mar 22 2011 - 10:42:49 CDT

I have been running equilibration of a single-chain protein complexed with one
ligand in a 14 Angstrom TIP3P box with PME boundary conditions. I use a 10A
cutoff and 8A switchdist. After 4ns, using 1fs timestep, I get this error and
cannot get around it by restarting.

I figure this error has to do with the boundary conditions used. I tried putting
in the option "splitPatch position" and using the last restart files, but it
keeps crashing at exactly the same timestep. I have also tried to remove the
wrapWater and wrapAll options but it does not help. What can I do to get around
this problem? Should I restart the simualtion from scratch using a different
cutoff distance, swtichdist, and solvent box size? Would disabling the wrapAll
and wrapWater options from start help? I have pasted my config file below:

set name 14A_box
set outputname ${name}_equil

set cbv1 77.937 ;# Box size in X dimension
set cbv2 90.367 ;# Box size in Y dimension
set cbv3 83.088 ;# Box size in Z dimension
set cox -2.118 ;# X coordinate for cell origin
set coy 1.681 ;# Y coordinate for cell origin
set coz -5.830 ;# Z coordinate for cell origin

coordinates ionized.pdb
structure ionized.psf
if {1} {
bincoordinates 14A_box_equil.restart.coor
binvelocities 14A_box_equil.restart.vel
extendedSystem 14A_box_equil.restart.xsc
firsttimestep 4034000

rigidBonds none

set T 310

set outputfreq 100
set dcdfreq 1000

paraTypeCharmm on
parameters /home/markus/MD_sim/common/opls_MD_mar_18_2011.prm

exclude scaled1-4
1-4scaling 1.0
switching on
cutoff 10
switchdist 8
pairlistdist 14

#temperature $T

langevin on
langevinDamping 0.5
langevinTemp $T
langevinHydrogen no

useGroupPressure no
useFlexibleCell no
useConstantArea no

langevinPiston on
langevinPistonTarget 1.01325
langevinPistonPeriod 100
langevinPistonDecay 50
langevinPistonTemp $T

cellBasisVector1 $cbv1 0. 0.
cellBasisVector2 0. $cbv2 0.
cellBasisVector3 0. 0. $cbv3
cellOrigin $cox $coy $coz

wrapWater on
wrapAll on

PME yes
PMEGridSpacing 1.0

stepspercycle 10
nonbondedFreq 1
fullElectFrequency 2

outputname $outputname
outputEnergies $outputfreq
outputPressure $outputfreq
outputTiming $dcdfreq
binaryoutput no
DCDfile $outputname.dcd
DCDfreq $dcdfreq
DCDUnitCell yes
binaryrestart yes
restartname $outputname.restart
restartfreq $dcdfreq
XSTfile $outputname.xst
XSTfreq $dcdfreq

timestep 1.0
run 20000000

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 05:23:42 CST