[gmx-users] npt.tpr protein-lig

[Christian Blau]

Hi Iman,


Great that you checked the notes!

The oscillatory period is estimated only from the masses and the force constant for the bond, see 
https://socratic.org/questions/frequency-of-vibration-of-two-masses-connected-by-a-spring ; note that the oscillation 
period = 2pi/\omega. To be on the conservative save side with the leap-frog algorithm, this then may not exceed 10 times 
your time-step to prevent integration errors that yield very unphysical results.

In your case, it seems that the parameters you are using for UNL seem to have a very large force constant on the bond 
between the atom types for C5 and C6 in your UNL parameters. One first step would be to re-check that the parameters you 
are using are fine and that you don't have an energy unit conversion error (note that GROMACS uses SI units) or that a 
Carbon atom type does not have the correct mass.

If you are reasonably convinced that your parameters are okay, reducing the timestep to 1fs from the 2fs that you have 
set currently (halving the performance of your run) might be an option, using constraints = all-bonds another. In 
general the latter is less preferable, unless your force-field was parametrized with this.

For temperature coupling you should use v-rescale instead of Berendsen in production runs, but that shouldn't affect the 
warning.

Best,

Christian

On 2019-12-02 13:02, Iman Katouzian wrote:
> Good day,
>
> I have generated npt.tpr for my protein ligand and changed my npt.mdp file
> according to the suggestions by GROMACS for using Brendsen and h-bonds for
> constraints finally I encounter this NOTE which I think will cause further
> problems in my md run :
>   NOTE 1 [file topol.top, line 18260]:
>    The bond in molecule-type UNL between atoms 6 C5 and 7 C6 has an
>    estimated oscillational period of 1.8e-02 ps, which is less than 10 times
>    the time step of 2.0e-03 ps.
>    Maybe you forgot to change the constraints mdp option.
>
> Number of degrees of freedom in T-Coupling group Protein_UNL is 4960.82
> Number of degrees of freedom in T-Coupling group Water_and_ions is 77850.18
> Determining Verlet buffer for a tolerance of 0.005 kJ/mol/ps at 300 K
> Calculated rlist for 1x1 atom pair-list as 1.034 nm, buffer size 0.034 nm
> Set rlist, assuming 4x4 atom pair-list, to 1.000 nm, buffer size 0.000 nm
>
> I would appreciate it if someone tells me what I need to do with my system
> I think it may be related to the time step of my simulaiton and frames as I
> searched the similar issues.
>
> thanks
>