Centering a protein in MD production run

User discussions
1

Hi gmx users,

I have a simulation of a system composed by a surface and a protein running for 80 ns in the production stage. This production stage is divided in four blocks of 20 ns, but I am facing a pbc problem with my protein in the second block since it is “getting out of the box”, I know this is due to the periodic boundary conditions, but I would like to know if there would be any problem in continuing running the missing stages this way and at the end of the final block just recenter the trajectory with gmx trjconv or it would be better to recenter the protein in my system before I continue with the others production blocks?

Please guide me.

2

There is no need to manipulate any files during simulations. The calculation of forces does not depend on our visual convenience of having a “centered” protein (which does not actually exist in an infinite system). Centering is only really relevant for visualization and some analyses.

3

Thank you very much for tour helpful response.

4

Dear Jalemkul,

Concerning your response. In which kind of analysis could be relevant to center with gmx trjconv?

5

Any analysis that depends on absolute coordinates and utilities that are not PBC-aware. RMSD is a big one, searching for molecules within a certain radius of the protein (if not PBC-aware), etc.

6

Thank you very much.