[gmx-users] Coulomb (SR) problem
David van der Spoel
spoel at xray.bmc.uu.se
Fri Jun 11 12:41:26 CEST 2010
On 2010-06-11 12.13, Thanasis Koukoulas wrote:
> hello again
>
> I had a problem with a sim of some molecules in vacuum and a gmx-user
> told me to do a simulation only in one molecule just to see what gromacs
> calculates and what does not. My problem concerned the sum of all
> energies in the system. After i did the sim with one molecule i found
> out that either gromacs either me calculate the Coulomb SR(short range)
> energy wrong. so for step 0 of this simulation I had this result:
>
> Started mdrun on node 0 Fri Jun 11 12:52:33 2010
>
> Step Time Lambda
> 0 0.00000 0.00000
>
> Grid: 3 x 3 x 3 cells
> Long Range LJ corr.: <C6> 3.8522e-03
> Long Range LJ corr.: Epot -0.00732091, Pres: -0.00379895, Vir: 0.00732091
> Energies (kJ/mol)
> Angle Ryckaert-Bell. LJ-14 Coulomb-14 LJ (SR)
> 5.70627e+00 8.37083e+00 6.80805e-03 1.72679e+02 -3.25801e+00
> Disper. corr. Coulomb (SR) Coul. recip. Potential Kinetic En.
> -7.32091e-03 -2.93587e+01 -1.21273e+02 3.28661e+01 3.94838e+01
> Total Energy Temperature Pressure (bar) Cons. rmsd ()
> 7.23499e+01 3.65291e+02 4.25496e+01 3.53180e-07
>
> so as you can see the value of Coulomb (SR) is
> -2.93587e+01 and my calculation is
> -1302.68
You should turn off PME and set the cutoffs to zero in this case.
>
> I believe that even i have done wrong calculations the value that
> gromacs finds is not correct.
> I found this value from the formula sum(f*qi*qj/rij) for every atom pair
> like
> 1-2,1-3,(not the 1-4),1-5....etc 2-3,2-4(not 1-2 because it is already
> calculated) ....etc ...3-.....etc......... inside the rcoulomb (the
> whole molecule)
> where f is the electric conversion factor f=138,935kJmol-1*nm*e-2
> IS COULOMB (SR) calculated that way by hand??? i an using PME btw
>
> i have not done arithmetic mistake i am sure for that (at least :) )
> Please consider that:
> 1) I dont use ffXXX.itp libraries from gromacs i have my own
> 2) I have tried every possible way to do this simulation for the past 3
> months (its my problem i know that !!! :) )
> 3) I have calculated the coulomb 1-4 with success and great accuracy
> 4) I have also caclulated the LJ (SR) with success (irrelevant!!!) :)
> 5) i am using gromacs 4.05
>
> Please if you have smth to advise me about this simulation except
> committing suicide of course :) please feel free to do it
>
> Thanks in advance
>
> especially Mark Abraham who helped my at first place
>
>
--
David van der Spoel, Ph.D., Professor of Biology
Dept. of Cell & Molec. Biol., Uppsala University.
Box 596, 75124 Uppsala, Sweden. Phone: +46184714205.
spoel at xray.bmc.uu.se http://folding.bmc.uu.se
More information about the gromacs.org_gmx-users
mailing list