Dear all,
I am running eABF simulations using CHARMM36 additive FF in NAMD for a few systems with water. The relevant input parameters are added below.
cutoff 12.0;
switchdist 10.0;
pairlistdist 16.0;
stepspercycle 20;
pairlistsPerCycle 2;
timestep 2.0; # fs/step
numsteps 50000000; # 1000ns
rigidBonds water #SHAKE_H-bonds
rigidTolerance 0.00001;
nonbondedFreq 1;
fullElectFrequency 1;
colvars on;
colvarsConfig colvars_abf_ext.in;
For a particular system, the error comes up:
ERROR: Atoms moving too fast; simulation has become unstable (1 atoms on patch 5 pe 69).
FATAL ERROR: Exiting prematurely
Even after varying different parameters, setting margin and starting from a different set of coordinates and minimizing enough, the error continues to occur. Since the atom in the picture is always H atom connected to C, I tried setting the rigid bonds to all instead of water, and the simulation has started running at a reduced timestep of 1 fs.
I want to ensure it is acceptable to set rigidBonds to all instead of water for eABF simulations. Any insights would be highly appreciated.
Thanks
Chythra
Don't have any idea, but still wanna get clarified. Have you tried changing other parameters ? Specially fullElectFrequency & rigidTolerance? Also setting rigidbonds to all might not be a suitable option I guess. Following...
Hello
This error is related to your modeling or minimization step. and also check your .pdb file xyz of all atom, if some directions are same you should change them manually
Chythra J N
Mohammad Ali Dashti Thank you for your reply.
I checked the system setup and the classical MD simulation using CHARMM36 ran perfectly well for microseconds. I also performed the minimization in CHARMM and NAMD before the actual eABF run. The .pdb file also doesn't have any atoms overlapping.
Even the simulation with rigidBonds set to all terminated with an error 'Constraint failure in RATTLE algorithm for atom 14! Constraint failure; simulation has become unstable' after running for 3/4th of the simulation time.
Is there any other possible solution to rectify this issue?
Chythra J N sometimes your time step could be changed , using lower time step (change 2fs to 1fs)
and also other trick that you can use is ,using high temp. for short time EM .
- Badges
- Science topic