Protein with Ion cofactors

User discussions
,
1

GROMACS version: 2024.1
GROMACS modification: No

Hello everyone,
I’m trying to simulate a protein dimer which has 3 calcium ions at a specific linker site. I can easily place the calciums where I want them in my PDB file and run all steps of the simulations. However, when I then use trjconv and remove pbc to center and rotate I can see the calcium ions “jump out” of the protein through short timepoints of the simulations (~200 ns total). I don’t think its a stability issue because they aren’t slowly leaving the protein but they rather blink in and out of it throughout the course of the trajectory.
The rest of the protein holds together fine but the calciums seem to have some problems. I’m wondering whether this could be a periodic boundary condition issue or something else. Is there a way to define ion cofactors as sort of part of the protein or to restrain their position with respect to the protein? I’d appreciate any guidance if people have dealt with similar systems.

2

It’s almost certainly an issue with jumps over the periodic boundary. It’s only a “problem” with the visualization - it does not affect the simulation. The -pbc nojump option would keep the calcium ions from jumping, but it would also results in, e.g., water molecules slowly diffusing away - seemingly into an empty space.

You could also try centering the trajectory on the ions (if they are a separate selection group) and then see if one of, or a combination of, the -pbc options would give you a suitable trajectory.