Lincs warning during NVT run at different Temp and Pressure

User discussions
1

GROMACS version: gmx mdrun, version 2019.3
GROMACS modification: No

I did energy minimization for 1 structure (asphaltene+pentane+pvc polymer), then I changed temperature and pressure (made 4 cases), and run NVT mdrun for 4 cases. I received lincs warning for 1 case, rest of the cases are run smoothly and now waiting for MD run. Here is the problem:

Step 7424, time 14.848 (ps) LINCS WARNING
relative constraint deviation after LINCS:
rms 24.123726, max 6325.152344 (between atoms 2181 and 2178) bonds that rotated more than 30 degrees:
atom 1 atom 2 angle previous, current, constraint length
2177 2175 90.0 0.1080 0.1273 0.1080
2181 2178 56.4 0.2356 683.2245 0.1080
2184 2180 90.0 0.2269 0.1296 0.1090
2185 2180 30.9 0.1375 0.1067 0.1090
2187 2182 90.0 0.1494 1.3018 0.1080
2188 2183 90.0 0.3104 0.3922 0.1090
2189 2183 90.0 0.1154 2.3238 0.1090
2196 2192 90.0 0.8155 1.0044 0.1090
2199 2194 90.0 0.1265 0.1448 0.1090
2200 2197 68.2 0.1465 0.1010 0.1010

Wrote pdb files with previous and current coordinates
/var/spool/slurmd/job13808560/slurm_script: line 7: 11355 Segmentation fault
(core dumped) gmx mdrun -v -deffnm nvt

Would you please tell me, why this lincs warning arise for 1 of the cases? How I solve this issue?
Thank you.

2

Please provide the contents of your full .mdp file. Note that pressure is irrelevant in an NVT simulation, so anything regarding ref-p will be ignored in the absence of a barostat.

3

Error comes from NVT-320K file. NVT 300K is good.

nvt-300K.mdp (1.7 KB) nvt-320K.mdp (1.7 KB)

4 
rlist       = 1.0       ; short-range neighborlist cutoff (in nm)
rcoulomb    = 1.4       ; short-range electrostatic cutoff (in nm)
rvdw        = 1.4       ; short-range van der Waals cutoff (in nm)

This part is a little odd, rlist should not be shorter than the nonbonded cutoffs. But that’s probably not the cause of the instability.

Do all simulations start from identical minimized coordinates? Does the 320 K run work if you reduce dt to e.g. 1 fs?

5

Do all simulations start from identical minimized coordinates — Yes

Does the 320 K run work if you reduce dt to e.g. 1 fs ---- I will check and let you know.

Thank you

6

Hi Justin

Both files were running and receive output without any warning/ error. I am using the Compute Canada cluster for my work. I am totally unsure why I faced this problem last week.

However, thank you for your support.