[gmx-users] extract subsystem energy: problem with mdrun -rerun and LJ(SR)

[Mark Abraham]

On 24/03/2010 7:34 PM, Andrea Minoia wrote:
> Dear all,
> I have performed a MD of a polymer chain interacting on a carbon
> nanotube using gromacs 4.0.7. I need to evaluate the potential energy of
> the polymer alone, during the dynamic, so to compare with its potential
> energy in vacuo. During the dynamic, the CNT was frozen, and I used
> exclusions groups to avoid to calculate CNT CNT non bonding interactions
> (I assume the bonds and angles potentials are computed anyway).
>
> Looking at the mailing list, I have used the following method:
>
> 1) use make_ndx to create a group system containing only the atoms for
> the polymer
>
> 2) Create a new simulation file:    tpbconv  -s original.tpr -n
> index.ndx -o rerun.tpr
> 3) adjust the trajectory according to the new simulation file:  trjconv
> -f totaltrj.xtc -n index.ndx -s rerun.tpr -o rerun.xtc
> 4) analyze the trajectory: mdrun -s rerun.tps -rerun rerun.xtc
>
> everything is performed without errors, but, when I use g_energy to
> evaluate the energy of the polymer, I get something like:
>
> Statistics over 15000001 steps [ 0.0000 thru 30000.0020 ps ], 7 data sets
> All averages are over 15001 frames
>
> Energy                      Average       RMSD     Fluct.      Drift
> Tot-Drift
> ------------------------------
> -------------------------------------------------
> Angle                       1630.18    61.5356    61.0703 -0.000872146
> -26.1644
> Ryckaert-Bell.              233.049    45.1358    34.5826 -0.00334919
> -100.476
> LJ-14                       669.907    20.0797    19.9169
> -0.000294607    -8.8382
> Coulomb-14                  637.187    16.2921    16.1707 0.000229164
> 6.87492
> LJ (SR)                           0          0          0
> 0          0
> Coulomb (SR)                      0          0          0
> 0          0
> Potential                   3170.33    84.0654    75.4238 -0.00428678
> -128.603
>
>
> The question is: why the short range lennard-jones and short range
> coulombic interactions are zero? those are the usual option from 1 to 8:
>    2  Angle            3  Ryckaert-Bell.   4  LJ-14   5
> Coulomb-14       6  LJ-(SR)          7  Coulomb-(SR)     8  Potential
>
> They are present for the orginal dynamic on CNT as well as in the
> dynamic in vacuo.
> hope in some answers. Thanks

In the absence of information about your ensemble (from your .mdp file 
or a description, hint hint) I'd guess it's because you've excluded all 
the non-bonded interactions for which you had non-zero parameters defined.

Mark