Metadynamics of the same systems are different

[Michaela]

Dear all,

I ran two metadynamics (metad) with the completely same system, however each has its own equilibration phase. The point is that each system has the same setup for metadynamics in the plumed.dat file, the same simulation time, the same molecules (protein around ribonucleic acid), but the FES file are different, so the final metadynamics looks different. Is it possible that we use wrong parameters for plumed? E.g. for the Gaussians?

Is there any proved approach how to test that our metadynamics makes sense? We are just curious why two runs of the same system behaves visibly different and want to be sure all of the setups are good.

Thank you for any help and suggestions.

[Massimiliano Bonomi]

Pretty normal if one or both of your simulations are not converged.
Check our tutorial to learn how to assess convergence (and calculate error bars):

https://www.plumed.org/doc-v2.7/user-doc/html/masterclass-21-4.html

Max

> --
> You received this message because you are subscribed to the Google Groups "PLUMED users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/6f5c009d-5313-4582-af2e-32152f970b79n%40googlegroups.com.

[Michaela]

Thanks for the particular tutorial. I found the jupyter notebook as well (https://github.com/plumed/masterclass-21-4/blob/main/notebooks/solution.ipynb).

I tried to plot free energy as function of time as it is in the cell No. 9, however I had to convert fes.dat with pandas "read_fwf" because I currently work on Windows (so I had to delete header of the fes files). The dataframe looks good. When I plot file.free vs cmap, I didnt get the expected curves but rather something looking like sticks:

Even plotting one fes.dat looks like this. I attached one fes.dat file and my jupyter notebook. I'm probably missing some important information or aspect of fes files, and I'd like to ask for any suggestions how to get expected curves when plotting such graphs. Thank you.

[Massimiliano Bonomi]

Hi,

I am not sure I understand what you are doing.
This is a free-energy surface as a function of 2 collective variables.
If you indeed plot the first, second, and third column of your fes file, looks ok to me

Max

> On Feb 2, 2022, at 16:52, Michaela <micha.c...@gmail.com> wrote:
>
> Thanks for the particular tutorial. I found the jupyter notebook as well (https://github.com/plumed/masterclass-21-4/blob/main/notebooks/solution.ipynb).
> I tried to plot free energy as function of time as it is in the cell No. 9, however I had to convert fes.dat with pandas "read_fwf" because I currently work on Windows (so I had to delete header of the fes files). The dataframe looks good. When I plot file.free vs cmap, I didnt get the expected curves but rather something looking like sticks:
>

> <fes-bad.png>
> Even plotting one fes.dat looks like this. I attached one fes.dat file and my jupyter notebook. I'm probably missing some important information or aspect of fes files, and I'd like to ask for any suggestions how to get expected curves when plotting such graphs. Thank you.
>
>
> On Tuesday, 1 February 2022 at 20:39:16 UTC+1 Max wrote:
> Pretty normal if one or both of your simulations are not converged.
> Check our tutorial to learn how to assess convergence (and calculate error bars):
>
> https://www.plumed.org/doc-v2.7/user-doc/html/masterclass-21-4.html
>
> Max
>
> > On Feb 1, 2022, at 20:36, Michaela <micha.c...@gmail.com> wrote:
> >
> > Dear all,
> >
> > I ran two metadynamics (metad) with the completely same system, however each has its own equilibration phase. The point is that each system has the same setup for metadynamics in the plumed.dat file, the same simulation time, the same molecules (protein around ribonucleic acid), but the FES file are different, so the final metadynamics looks different. Is it possible that we use wrong parameters for plumed? E.g. for the Gaussians?
> > Is there any proved approach how to test that our metadynamics makes sense? We are just curious why two runs of the same system behaves visibly different and want to be sure all of the setups are good.
> > Thank you for any help and suggestions.
> >
> > --
> > You received this message because you are subscribed to the Google Groups "PLUMED users" group.
> > To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users...@googlegroups.com.
> > To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/6f5c009d-5313-4582-af2e-32152f970b79n%40googlegroups.com.
>
>
> --
> You received this message because you are subscribed to the Google Groups "PLUMED users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/baa1a46a-29f3-48e5-a599-e546975da4d8n%40googlegroups.com.
> <fes-bad.png><test.ipynb><fes_95-Copy1.dat>