unstable simulation with AMBER ff in AMBER!

From: Lina Nilsson (linan_at_u.washington.edu)
Date: Wed Apr 13 2005 - 04:05:36 CDT

Dear all,
I need to use AMBER ff's for a simulation, but I am unable to get a
stable MD system in NAMD from .crd and .top files created with xleap
(tried both a protein I have previously simulated using charmm and also
a two amino acid 'test case.') Do I need to add further minimzation
steps when using AMBER ff's? Any other ideas??

DETAILS:
My periodic boxes (TIP3 water and protein) blow up as soon as the
temperature is turned on (the energies, volumes etc seem to be
fine/plataeu during mimimization in NAMD, as far as I can tell). But as
I start thermalizing, the system volume shrinks rapidly, only to then
start increasing (as do temps and energies) and then 'blowing up' as
atoms start moving too fast for the simulation to handle.

The NAMD base scripts I use work fine for charmm ff based simulations.
The densities I get from the xleap prep are low (<0.8 g/cc)compared to
what I get in charmm (reducing the closeness parameter in xleap to 0.5
doesn't change the density much). May this be causing instabilities?
Turning on RigidBonds in NAMD reduces the problem (ie slows down the
temp/vel increase), but does not get rid of it, for what that is worth.
I am completely stuck on this and ANY input is much appreciated!

Lina

TWO AMINO ACID TEST CASE

XLEAP:

xleap -s -f $AMBERHOME/dat/leap/cmd/leaprc.ff # tried ff03 and then ff94

filename = loadpdb PDB.pdb
solvatebox filename TIP3PBOX 16
charge filename # --> no charge
check filename
# --> no errors. "Unit is ok" Some hydrogens are too close to each other
saveamberparm filename filename.top filename.crd
savepdb filename filename.pdb

#Other than the water density, the system looks fine in VMD (no missing
atoms etc). There are no warnings / obvious errors in my xleap log file.

THEN MINIMIZING IN NAMD in 3 steps (first water only, then holding
protein backbone still, then all atom minimization):

cwd ./
amber on
readexclusions no # also tried reading exclusions, although this
shouldn't matter since the parameters should all be defined below.
parmfile ff94_xxx.top
coordinates ff94_xxx.pdb

restartname xxx.rst
restartfreq 100
binaryrestart yes
outputname xxx
binaryoutput no
DCDfile xxx.dcd
DCDfreq 100
outputEnergies 1
outputMomenta 1
outputPressure 1
XSTfile xxx.xst
XSTfreq 100

numsteps 2000
timestep 1.0

fixedAtoms on
fixedAtomsForces on
fixedAtomsFile fixedAtoms.pdb
fixedAtomsCol B

stepspercycle 4
nonbondedFreq 1
cutoff 12.0
switching on
switchdist 10.0
pairlistdist 13.0
exclude scaled1-4

1-4scaling 0.833333
fullElectFrequency 4
PME yes
PMETolerance 0.000001
PMEInterpOrder 4
PMEGridSizeX 36
PMEGridSizeY 36
PMEGridSizeZ 36

scnb 2
temperature 0
dielectric 1.0

cellBasisVector1 34.572 0 0
cellBasisVector2 0 33.2880 0
cellBasisVector3 0 0 33.2840
cellOrigin -0.1518 0.1325 -0.0739
wrapWater on

constraints off
minimization on

--> output *seems* ok...

FINALLY, THERMALIZING:

cwd ./
amber on
readexclusions no
parmfile ff94_xxx.top
coordinates minoutput.coor
extendedSystem minoutput.xsc

restartname xxx.rst
restartfreq 100
binaryrestart yes
outputname xxx
binaryoutput no
DCDfile xxx.dcd
DCDfreq 100
outputEnergies 100
outputMomenta 100
outputPressure 100
XSTfile xxx.xst
XSTfreq 100

numsteps 40000
timestep 1.0

fixedAtoms on
fixedAtomsForces on
fixedAtomsFile fixedAtoms.pdb
fixedAtomsCol B

stepspercycle 4
nonbondedFreq 1
cutoff 12.0
switching on
switchdist 10.0
pairlistdist 13.0
exclude scaled1-4

1-4scaling 0.833333
fullElectFrequency 4
PME yes
PMETolerance 0.000001
PMEInterpOrder 4
PMEGridSizeX 36
PMEGridSizeY 36
PMEGridSizeZ 36

scnb 2
temperature 0
dielectric 1.0

reassignFreq 1000
reassignTemp 10.0
reassignIncr 10.0
reassignHold 310.0

useGroupPressure no
useFlexibleCell no
BerendsenPressure on
BerendsenPressureTarget 1.01325
BerendsenPressureCompressibility 0.000045

BerendsenPressureRelaxationTime 1000
BerendsenPressureFreq 4
wrapWater on
constraints off

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:39:20 CST