[gmx-users] Maximal nrexcl

[cynthiahong1983]

Thanks Justin.
I did not use energy group or energy group exclusions, because my polymers carry charges. Energy-group methods seem to be unable to calculate different contributions into electrostatic energy in reciprocal space. 
Can grompp be run parallelly so that the memory load can be divided onto nodes ? 
Thank you.Bingster------------------------------------------------------------------




On 1/20/14, 8:17 PM, cynthiahong1983 wrote:
> Dear All,
>
> Does anyone know whether there is a maximal limit for nrexcl?
>
> I tried to calculate the intermolecular energy between polymeric molecules through "mdrun -rerun" and a large "nrexcl". But when the molecule becomes very long e.g. having ~ 20 bonds in the backbone and I set nrexcl = 20 , grompp beccame very slowly and just terminated in error like "4210 terminated" (the number changed each time when it failed.)
>
> Someone ever met with similar situations?
>

The only real limitation is available memory.  For this case, you'd need a lot 
of it.  The other option (depending on how many polymers you have) would be to 
assign each one to individual energygrps in the .mdp and .ndx files.

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 601
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441

==================================================