- Posts: 90
Ternary lipid simulation fails after 2.14 microsec
- Rajat Desikan
-
Topic Author
- Offline
- Expert Boarder
Less
More
10 years 5 months ago #2810
by Rajat Desikan
Ternary lipid simulation fails after 2.14 microsec was created by Rajat Desikan
Hi all,
I am simulating a DMPC:DOPC:CHOL ternary mixture using elnedyn22.ff (All ring masses set to 72 amu in martini.itp but changed back to 45 in cholesterol.itp). The simulation was run in NPT using a 20 fs timestep and -rdd 1.6
It fails after approximately 2.14 microsec with the below error message (basically my system is blowing up). The system consists of about 2000 lipids+water and was run on 128 processors. Can anyone help me identify what went wrong and how to correct it? Thank you.
P.S. I intend to run this systems and possibly membrane-protein systems much longer.
Step Time Lambda
107133000 2142660.00000 0.00000
Energies (kJ/mol)
Bond G96Angle Improper Dih. LJ (SR) Coulomb (SR)
4.35718e+04 1.07417e+04 2.43858e+04 -1.08448e+06 -4.82380e+03
Potential Kinetic En. Total Energy Temperature Pressure (bar)
-1.01060e+06 1.76903e+05 -8.33702e+05 3.10779e+02 1.11272e+01
Constr. rmsd
7.59236e-04
Not all bonded interactions have been properly assigned to the domain decomposition cells
A list of missing interactions:
Bond of 20256 missing 3
Molecule type 'CHOL'
the first 10 missing interactions, except for exclusions:
Bond atoms 5 6 global 8665 8666
Molecule type 'DOPC'
the first 10 missing interactions, except for exclusions:
Bond atoms 5 6 global 18103 18104
Bond atoms 6 7 global 18104 18105
Program mdrun_mpi, VERSION 4.6.3
Source code file: /home/rajat/softback/gromacs-4.6.3/src/mdlib/domdec_top.c, line: 393
Fatal error:
3 of the 37444 bonded interactions could not be calculated because some atoms involved moved further apart than the multi-body cut-off distance (1.6 nm) or the two-body cut-off distance (1.6 nm), see option -rdd, for pairs and tabulated bonds also see option -ddcheck
For more information and tips for troubleshooting, please check the GROMACS
website at www.gromacs.org/Documentation/Errors
[/b]
Here is my .mdp
integrator = md
dt = 0.02
nsteps = 125000000 ; 2.5 microsec
nstcomm = 10
comm-grps = DMPC_DOPC_CHOL W
nstxout = 10000000
nstvout = 10000000
nstfout = 10000000
nstlog = 1000
nstenergy = 1000
nstxtcout = 10000 ;every 200 ps
xtc_precision = 1000
energygrps = DMPC DOPC CHOL W
nstlist = 10
ns_type = grid
pbc = xyz
rlist = 1.4
coulombtype = Shift ;Reaction_field (for use with Verlet-pairlist)
rcoulomb_switch = 0.0
rcoulomb = 1.2
epsilon_r = 15 ; 2.5 (with polarizable water)
vdw_type = Shift ;cutoff (for use with Verlet-pairlist)
rvdw_switch = 0.9
rvdw = 1.2 ;1.1 (for use with Verlet-pairlist)
tcoupl = v-rescale
tc-grps = DMPC_DOPC_CHOL W
tau_t = 1.0 1.0
ref_t = 310 310
Pcoupl = parrinello-rahman
Pcoupltype = semiisotropic
tau_p = 12.0 12.0 ;parrinello-rahman is more stable with larger tau-p, DdJ, 20130422
compressibility = 1e-5 1e-5 ;
ref_p = 1.0 1.0
gen_vel = no
constraints = none
constraint_algorithm = Lincs
lincs_order = 4
lincs_warnangle = 30
continuation = yes
I am simulating a DMPC:DOPC:CHOL ternary mixture using elnedyn22.ff (All ring masses set to 72 amu in martini.itp but changed back to 45 in cholesterol.itp). The simulation was run in NPT using a 20 fs timestep and -rdd 1.6
It fails after approximately 2.14 microsec with the below error message (basically my system is blowing up). The system consists of about 2000 lipids+water and was run on 128 processors. Can anyone help me identify what went wrong and how to correct it? Thank you.
P.S. I intend to run this systems and possibly membrane-protein systems much longer.
Step Time Lambda
107133000 2142660.00000 0.00000
Energies (kJ/mol)
Bond G96Angle Improper Dih. LJ (SR) Coulomb (SR)
4.35718e+04 1.07417e+04 2.43858e+04 -1.08448e+06 -4.82380e+03
Potential Kinetic En. Total Energy Temperature Pressure (bar)
-1.01060e+06 1.76903e+05 -8.33702e+05 3.10779e+02 1.11272e+01
Constr. rmsd
7.59236e-04
Not all bonded interactions have been properly assigned to the domain decomposition cells
A list of missing interactions:
Bond of 20256 missing 3
Molecule type 'CHOL'
the first 10 missing interactions, except for exclusions:
Bond atoms 5 6 global 8665 8666
Molecule type 'DOPC'
the first 10 missing interactions, except for exclusions:
Bond atoms 5 6 global 18103 18104
Bond atoms 6 7 global 18104 18105
Program mdrun_mpi, VERSION 4.6.3
Source code file: /home/rajat/softback/gromacs-4.6.3/src/mdlib/domdec_top.c, line: 393
Fatal error:
3 of the 37444 bonded interactions could not be calculated because some atoms involved moved further apart than the multi-body cut-off distance (1.6 nm) or the two-body cut-off distance (1.6 nm), see option -rdd, for pairs and tabulated bonds also see option -ddcheck
For more information and tips for troubleshooting, please check the GROMACS
website at www.gromacs.org/Documentation/Errors
[/b]
Here is my .mdp
integrator = md
dt = 0.02
nsteps = 125000000 ; 2.5 microsec
nstcomm = 10
comm-grps = DMPC_DOPC_CHOL W
nstxout = 10000000
nstvout = 10000000
nstfout = 10000000
nstlog = 1000
nstenergy = 1000
nstxtcout = 10000 ;every 200 ps
xtc_precision = 1000
energygrps = DMPC DOPC CHOL W
nstlist = 10
ns_type = grid
pbc = xyz
rlist = 1.4
coulombtype = Shift ;Reaction_field (for use with Verlet-pairlist)
rcoulomb_switch = 0.0
rcoulomb = 1.2
epsilon_r = 15 ; 2.5 (with polarizable water)
vdw_type = Shift ;cutoff (for use with Verlet-pairlist)
rvdw_switch = 0.9
rvdw = 1.2 ;1.1 (for use with Verlet-pairlist)
tcoupl = v-rescale
tc-grps = DMPC_DOPC_CHOL W
tau_t = 1.0 1.0
ref_t = 310 310
Pcoupl = parrinello-rahman
Pcoupltype = semiisotropic
tau_p = 12.0 12.0 ;parrinello-rahman is more stable with larger tau-p, DdJ, 20130422
compressibility = 1e-5 1e-5 ;
ref_p = 1.0 1.0
gen_vel = no
constraints = none
constraint_algorithm = Lincs
lincs_order = 4
lincs_warnangle = 30
continuation = yes
Please Log in or Create an account to join the conversation.
- Rajat Desikan
-
Topic Author
- Offline
- Expert Boarder
Less
More
- Posts: 90
10 years 5 months ago #2851
by Rajat Desikan
Replied by Rajat Desikan on topic Ternary lipid simulation fails after 2.14 microsec
Hi All,
Any suggestions? I am stuck here. I tried re-running it with a different starting velocity. It failed in the vicinity of 2 microsec with the same error...
Thanks,
Any suggestions? I am stuck here. I tried re-running it with a different starting velocity. It failed in the vicinity of 2 microsec with the same error...
Thanks,
Please Log in or Create an account to join the conversation.
- Clement
-
- Offline
- Admin
Less
More
- Posts: 211
10 years 5 months ago - 10 years 5 months ago #2852
by Clement
Replied by Clement on topic Ternary lipid simulation fails after 2.14 microsec
It happens once in a while... Probably a cholesterol turned crazy. Just restart the dynamics from the checkpoint file (*.cpt or *_prev.cpt). You won't lose your 2.14 us.
Last edit: 10 years 5 months ago by Clement.
Please Log in or Create an account to join the conversation.
- xavier
-
- Offline
- Admin
Less
More
- Posts: 416
10 years 5 months ago #2896
by xavier
Replied by xavier on topic Ternary lipid simulation fails after 2.14 microsec
As Clement said this is indeed happening as the cholesterol tends to be unstable at dt=20 fs. You can indeed restart your simulation using the cpt and it should go fine.
For your next simulation you may want to copy the mdout.mdp into your original mdp and modify the nstcalcenergy/temperature/pressure. These are by default set to 100 (the nstcomm will be adjusted to these). This setup is basically fine for atomistic simulations but for a CG forcefield using a 20 fs time step it seems to add numerical instability. I generally use nstcalcenergy = 10.
For your next simulation you may want to copy the mdout.mdp into your original mdp and modify the nstcalcenergy/temperature/pressure. These are by default set to 100 (the nstcomm will be adjusted to these). This setup is basically fine for atomistic simulations but for a CG forcefield using a 20 fs time step it seems to add numerical instability. I generally use nstcalcenergy = 10.
Rajat Desikan wrote: Hi all,
I am simulating a DMPC:DOPC:CHOL ternary mixture using elnedyn22.ff (All ring masses set to 72 amu in martini.itp but changed back to 45 in cholesterol.itp). The simulation was run in NPT using a 20 fs timestep and -rdd 1.6
It fails after approximately 2.14 microsec with the below error message (basically my system is blowing up). The system consists of about 2000 lipids+water and was run on 128 processors. Can anyone help me identify what went wrong and how to correct it? Thank you.
P.S. I intend to run this systems and possibly membrane-protein systems much longer.
Step Time Lambda
107133000 2142660.00000 0.00000
Energies (kJ/mol)
Bond G96Angle Improper Dih. LJ (SR) Coulomb (SR)
4.35718e+04 1.07417e+04 2.43858e+04 -1.08448e+06 -4.82380e+03
Potential Kinetic En. Total Energy Temperature Pressure (bar)
-1.01060e+06 1.76903e+05 -8.33702e+05 3.10779e+02 1.11272e+01
Constr. rmsd
7.59236e-04
Not all bonded interactions have been properly assigned to the domain decomposition cells
A list of missing interactions:
Bond of 20256 missing 3
Molecule type 'CHOL'
the first 10 missing interactions, except for exclusions:
Bond atoms 5 6 global 8665 8666
Molecule type 'DOPC'
the first 10 missing interactions, except for exclusions:
Bond atoms 5 6 global 18103 18104
Bond atoms 6 7 global 18104 18105
Program mdrun_mpi, VERSION 4.6.3
Source code file: /home/rajat/softback/gromacs-4.6.3/src/mdlib/domdec_top.c, line: 393
Fatal error:
3 of the 37444 bonded interactions could not be calculated because some atoms involved moved further apart than the multi-body cut-off distance (1.6 nm) or the two-body cut-off distance (1.6 nm), see option -rdd, for pairs and tabulated bonds also see option -ddcheck
For more information and tips for troubleshooting, please check the GROMACS
website at www.gromacs.org/Documentation/Errors
[/b]
Here is my .mdp
integrator = md
dt = 0.02
nsteps = 125000000 ; 2.5 microsec
nstcomm = 10
comm-grps = DMPC_DOPC_CHOL W
nstxout = 10000000
nstvout = 10000000
nstfout = 10000000
nstlog = 1000
nstenergy = 1000
nstxtcout = 10000 ;every 200 ps
xtc_precision = 1000
energygrps = DMPC DOPC CHOL W
nstlist = 10
ns_type = grid
pbc = xyz
rlist = 1.4
coulombtype = Shift ;Reaction_field (for use with Verlet-pairlist)
rcoulomb_switch = 0.0
rcoulomb = 1.2
epsilon_r = 15 ; 2.5 (with polarizable water)
vdw_type = Shift ;cutoff (for use with Verlet-pairlist)
rvdw_switch = 0.9
rvdw = 1.2 ;1.1 (for use with Verlet-pairlist)
tcoupl = v-rescale
tc-grps = DMPC_DOPC_CHOL W
tau_t = 1.0 1.0
ref_t = 310 310
Pcoupl = parrinello-rahman
Pcoupltype = semiisotropic
tau_p = 12.0 12.0 ;parrinello-rahman is more stable with larger tau-p, DdJ, 20130422
compressibility = 1e-5 1e-5 ;
ref_p = 1.0 1.0
gen_vel = no
constraints = none
constraint_algorithm = Lincs
lincs_order = 4
lincs_warnangle = 30
continuation = yes
Please Log in or Create an account to join the conversation.
Time to create page: 0.098 seconds