Compare qm potentials with forcefield potential energy

User discussions
1

Hi all,

I researching two molecules; Chlorophyll and Histidine from LHCII in spinach. I isolated the two molecules and made an optimization scan, with different distances between the MG atom of Chlorophyll an and NE2 atom of Histidine, with quantum calculations. The result looks like “nice” Lennard jones potential.

Now my proposition is to get the same potential shape in Gromacs. For that, I introduced a new covalent bond between MG and NE2 and calculated the force constant from the quantum calculations. From the mdruns, I always took the potential energy. Now, the potentials from mdruns look similar to the quantum potential, however it still “hopping” around. My approach was to apply a position restraint to both atoms, so that the system can relax in energy minimization. This however, had no significant effect. My second approach was to execute a distance restrain in the new bond, which lowered the potential energy by two magnitudes, but the potential didn’t look very good. Does anyone here have a suggestion, how I could relax my system, to get a “nice” potential curve?

Thanks in advance

2

You mean, you want to compare the relaxed energy scan in QM to a series of relaxed energy minimizations in MM? If you want to consider the interaction energy between the two, and you can perform a rerun on your QM geometries (that is, you have a compatible topology), I’m not sure if you need to relax the geometries - the planar parts will be pretty much unaffected.

Here’s a tutorial I made for quickly calculating such interaction energies with Gromologist:

Note that for the comparison to make sense, you should run your QM scan with the counterpoise/BSSE correction - this way you won’t include the changes in internal residue energies on either level.

1 Like