Possible sexual error: the box size is too small

Continuing the discussion from There is an error in input file(S):

[quote=“kesistor, post:1, topic:7012, full:true”]
GROMACS version:2022
GROMACS modification: No

When I input the command line:
gmx grompp -f npt_umbrella.mdp -c conf207.gro -p topol.top -r conf207.gro -n index.ndx -o npt4.tpr

it will remind me:
There is an error in input file(S)

The details are as follows:
ERROR 1[file npt_umbrella.mdp]
When the maximum distance from a pull group referrence atom to other atoms in the group is larger than 0.5 times half the box size a centrally placed atom should be chosen as pbctom. Pull group 1 is larger than that and does not have a specific atom selected as reference atom.

But the key is that I can successfully run the command
gmx grompp -f npt_umbrella.mdp -c conf6.gro -p topol.top -r conf6.gro -n index.ndx -o npt0.tpr

The npt1~3 has been created. This means some parts are OK, the others are not.
What’s more, I got file npt_umbrella.npt from GROMACS tutorial, which means it shouldn’t be wrong theoretically.
-----------------------the second update-----------------------------------
The pull distance fills the entire box space when viewing molecular images in VMD.In my opinion, the error may be the unsuitable box size.

Without more information it’s difficult to say exactly what’s the problem. My guess would be that pull group 1 is more extended after the first three NPT simulation stages. Have you checked the trajectories?

The error itself is easy to address. Just do what it says. You can find more information in the manual: https://manual.gromacs.org/current/reference-manual/special/pulling.html#definition-of-the-center-of-mass. You need to select one centrally located atom to use as a reference when calculating the center of mass, taking periodic images into account.

As you say, in principle tutorials should work, but since GROMACS is under development you must make sure that the tutorial was developed for the version of GROMACS you are using.