Dear Amberites,
I've noticed what may be a problem running QM/MM calculations with the Amber22/Gaussian16 interface.
In one calculation I'm running, all the Gaussian QM steps are crashing and dumping a core. The QM log file ends with:
Leave Link 303 at Tue Apr 30 22:51:26 2024, MaxMem= 8589934592 cpu: 99.0 elap: 2.8
(Enter /apps/gaussian/16/avx2/g16/l401.exe)
Initial guess from the checkpoint file: "gau_job.chk"
Basis set data is not on the checkpoint file.
Error termination via Lnk1e in /apps/gaussian/16/avx2/g16/l401.exe at Tue Apr 30 22:51:27 2024.
Job cpu time: 0 days 0 hours 43 minutes 48.6 seconds.
Elapsed time: 0 days 0 hours 0 minutes 45.9 seconds.
File lengths (MBytes): RWF= 546 Int= 0 D2E= 0 Chk= 266 Scr= 1
This is caused because the "Guess=Read" directive is added to the Gaussian input, but for some reason there's no basis set information in the checkpoint file.
I suppose when that happes, one should expect the whole calculation to crash. However, somehow Amber keeps going. every MD step reports as successful:
Extract from the mdout file:
>>>>> Entered write_inpfile (qm2_extern_gau_module)
<<<<< Left write_inpfile (qm2_extern_gau_module)
Input file written successfully; calling Gaussian...
Gaussian execution success; Processing Gaussian results...
>>>>> Entered read_results (qm2_extern_gau_module)
<<<<< Left read_results (qm2_extern_gau_module)
>>>>> Entered print_results (qm2_extern_util_module)
<<<<< Left print_results (qm2_extern_util_module)
NSTEP = 0 TIME(PS) = 0.000 TEMP(K) = 0.00 PRESS = 0.0
Etot = 948088.5544 EKtot = 0.0000 EPtot = 948088.5544
BOND = 724.6589 ANGLE = 1917.2682 DIHED = 1337.2657
1-4 NB = 847.5750 1-4 EEL = 9564.9226 VDWAALS = 15737.1757
EELEC = -169077.2411 EHBOND = 0.0000 RESTRAINT = 0.0000
EXTERNESCF= 1086869.2158
CMAP = 167.7136
Ewald error estimate: 0.4605E-04
------------------------------------------------------------------------------
Has anyone noticed that before? I'm aware Amber is already on version 25, but 22 is the highest I have access to at the moment. I don't know if this has been seen/addressed in a more recent release.
I'd be happy to share my input files if anyone wants to look at that.
All the best,
Gustavo.
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Wed May 01 2024 - 08:30:02 PDT
