I checked the topology file that I generated for TIP4P water using LEAP
and the one ($AMBERHOME/test/tip4p/prmtop) in the Amber distribution.
The difference between them is the mass of the "EP". In my topology
file, which derived from the "frcmod.tip4pew" file, the mass of "EP" is
zero.
But mass of "EP" in the "prmtop" file is 3.00. I changed the mass of
"EP" in the "frcmod.tip4pew" file and regenerated the topology, sander
seems to run fine!
My question is why the mass of water is 21 (2 for 2H + 16 for O and
3 for EP) instead of 18? If not, how can I deal with this problem?
Evidently, sander does not like zero mass regardless of the type.
Thanks,
--Dat
David A. Case wrote:
>On Thu, Jun 10, 2004, Dat H. Nguyen wrote:
>
>
>>I tried to simulate a system of about 5400 TIP4P water molecules (only
>>water, not solute) using the parallel version of sander module in AMBER
>>8 and I use frcmod.tip4pew forcefield parameters. However, I got a core
>>dump with the following error:
>>
>>forrtl: error (73): floating divide by zero
>> 0: __FINI_00_remove_gp_range [0x3ff81a21488]
>> 1: __FINI_00_remove_gp_range [0x3ff81a2a200]
>> 2: __FINI_00_remove_gp_range [0x3ff800d5d10]
>> 3: rdparm2_ [_rdparm.f: 1584, 0x1200936b4]
>> 4: sander_ [_sander.f: 2496, 0x12005a538]
>> 5: multisander_ [_sander.f: 885, 0x1200592c0]
>> 6: main [for_main.c: 203, 0x12013d40c]
>> 7: __start [0x1200250c8]
>>/usr/local/bin/mpirun: 450484 Abort - core dumped
>>
>>
>
>There is a tip4p test case in the Amber distribution: you might try to figure
>out how your prmtop differs from the one there.
>
>...dac
>
>
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Tue Jun 15 2004 - 22:53:01 PDT
