Is a triclinic box suitable for simulating an elongated structure?

User discussions
1

GROMACS version: 2023.3
GROMACS modification: No

There is a protein that has an elongated system. If I place it in a triclinic box, this will allow me to optimize the size of the system (reduce the number of atoms). At the same time, as a result of fluctuations, there is a risk that the protein will turn and part of it will end up outside the box. For this reason, several questions arise:

  1. can a situation where part of the protein is outside the box affect the correctness of the calculations (such as rmsd, rmsf, gyration, etc)?
  2. in this case, it is advisable to use a smaller triclinic box, or it is better to place the protein in a larger dodecahedron box, which excludes the possibility of the end protein leaving the boundaries of the box during molecular dynamics fluctuations?
2

Leaving the box is never an issue. The actual, potential, issue is that an elongated molecule might interact stronger when it gets closer to its periodic image due to rotation.

You might be able to use angular center of mass motion removal to avoid this.