Oscillation period is less than time step in Charmm36ff

User discussions
1

GROMACS version:
GROMACS modification: Yes/No

System consist of protein and RNA minimized in the Charmm36ff. When running nvt and npt there is the following Note:

NOTE 1 [file topol.top, line 50]:
The bond in molecule-type RNA_chain_B between atoms 56 C2 and 57 O2 has
an estimated oscillational period of 1.9e-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.

It is necessary to add that in mdp file constraints were set to = h-bonds as required.
Does this note is critical for equilibration and further molecular dynamics? What may be a general approach to solve this problem? Does the change of time step to 1fs (dt =0.001) will be the right solution?

2

If you want to be on the safe side, you could set the time step to 1.5 fs. But this note is just to make you aware of a potentially risky setting. If the factor gets below 5 it will be a warning instead.

In your case you are very close to a factor 10 difference. I think you’ll be fine with a 2 fs time step, but I can’t guarantee anything.

3

Is it necessary to change tau_t and tau_p if I change the time step (df)?

I used two-stage equilibration befor md simulation and I‘m not sure if some other parameters should be changed (if they are correct at all):

Equilibration 1st stage: v-rescale, tau_t =0.1 ps (for 1-10 ns)
Equilibration 2nd stage: v-rescale, tau_t=0.1 ps, c-rescale, tau_p = 2.0 ps (for 10-100 ns)
MD: v-rescale, tau_t=0.1 ps, c-rescale, tau_p = 2.0 ps

I see that pressure and temperature are rather stable when using this parameters, but still not sure about them.

4

Within reasonable limits, the time constants for pressure and temperature coupling do not depend on the simulation time step.

In general, I would recommend slightly larger time constants, especially for the temperature coupling. See if What is the recommended value for tau-p and tau-t with GROMACS2023.0 when using PR and NH helps you.