GROMACS version: 2022
GROMACS modification: No
Here post your question
Hello,
I have an grompp warning about the atoms excluded during energy minimization. Please help me understand why it occured and if it requires adjustment of input or files.
I understand that Energy minimization determines the bonded and nonbonded interactions between neighbors and that with my basic mdp file I set the cut off for bonded interactions to 1.2nm.
when I ran my input the result produced excluded 3 atoms (bonded neighbors) from my protein water and ligands.
the warning :
The largest distance between excluded atoms is 1.160 nm, which is larger
than 90% of the cut-off distance. When excluded pairs go beyond the
cut-off distance, this leads to missing long-range corrections in the
forces and energies.
if the largest distance is still < the cut-off why are the atoms excluded from bonded neighbors?
Also any links to better improve my functional understanding of the parameters to set when performing energy minimization are appretiated.
You’ve probably fixed your problem by now, but just in case you are still puzzled: The message is telling you that you have 2 bonded neighbour atoms 1.16nm (11.6A) apart. This indicates that your system is set up wrong! Take a look in VMD and see if you can find the issue.
It looks like you’ve got a break in your protein chain between residues 263 and 269. Pdb2gmx used to handle that silently before 2022, but it now leads to an error downstream when you try and use the structure. You can fix the problem by building in the missing residues (in e.g. Coot), or you can fudge it by adding a TER record to your PDB file at the breakpoint (but this may lead to incorrect physical behaviour).
I’ve just hit this problem myself. I’m not sure whether this change in pdb2gmx was intentional or not. I need to do some more checks before asking the developers for help.