Unbiasing metadynamics MD data ran with colvars module

User discussions
1

GROMACS version: 2024.1
GROMACS modification: No

Hello,
I’m attempting to apply well-tempered metadynamics to a small disordered peptide using the colvars module. When analysing the data, since the simualtions has been biased using a number of CVs, I assume one should unbias the data before performing any analysis which is to be compared to standard brute force MD.
Any recommendations how this can be done? I’ve seen that PLUMED might contain useful methods, however I am unsure if they can be used due to the simulations not being ran with PLUMED.

Any help is appreciated.

2

Hi @Sveakso I think the answer highly depends on what the PLUMED analysis tool expects to read. If you want to have the equivalent of the HILLS file try the writeHillsTrajectory option.

If that’s not enough, can you describe what kind of analysis you plan to carry out after the simulation?

Hope that helps,
Giacomo

3

Hi @giacomo.fiorin,
Thank you so much for the reply! I was indeed looking for something such as the HILLS files and will try the option you presented. Can the output files from this be used in PLUMED?

On top of reconstruction of the free energy landscape, I am planning on calculating the X-ray scattering spectra which I have a method for. My curiosity is, in the case of metadynamics should/can one carry out an unbiasing of the tradjectory before such an analysis or is analysis simply performed like usual?

4

Yes :-)