Calculation of Interaction energies

User discussions
1

GROMACS version: 2021
GROMACS modification: No

Hello!

I have a system containing a membrane and two proteins. I would like to compute the interaction energies:

  • between the membrane and Protein 1

  • between the membrane and Protein 2

  • between Protein 1 and Protein 2

I am unsure about the best approach.

One option I considered is generating separate trajectories for each pair of components, rerunning the energy calculation, and comparing their potential energies.

Alternatively, I could define energy groups for the full system and extract:

  • LJ-SR interactions

  • Coul-SR interactions

However, I am concerned that this method only includes non-bonded interactions.

Could someone advise which approach is more appropriate and commonly used for estimating interaction energies in such systems?

Thank you in advance!

2

You do not need to define separate trajectories. Instead, you can rerun to recalculate energies for the relevant energy groups. These can be defined using an index (ndx) file, although this may be not even needed, depending on your topology. If you include the desired sets of energy-grps in the mdp file, generate a new tpr file and then run gmx mdrun -rerun with the new tpr file, you should be able to get what you want in the energy (edr) file.

A note of warning however from a previous question.

And also, I am assuming your groups are covalently bonded, so non-bonded interactions are all there is.

I hope this helps,

David

3

If you want to go beyond interaction energy, check out the recipe in this old paper (details in SI). https://pubs.acs.org/doi/10.1021/ct400404q

4

Thanks @daviddesancho and @dspoel

5

Just a quick check: when extracting interaction energies using gmx energy, the values are reported in kJ/mol, correct? There isn’t any case where they appear as kJ/mol but are actually in J/mol, right?