Grompp for polymer model at old Gromacs version (4.5.7)

Hi,

I have built an OPLS-AA model of Polymer (PMMA) at the site below.

I would like to run this model on Gromacs 4.5.7.
The reason for using this version is that it is customized.
The above site creates a pmma.pdb file and a pmma.itp file.
When I changed the the topology file as follows and ececuted the grompp, the error message was displayed.

topology.top

[ defaults ]
; nbfunc comb-rule gen-pairs fudgeLJ fudgeQQ
1 3 yes 0.5 0.5

#include “p1_raw.itp”
[ system ]
PMMA

[ molecules ]
LIG 10

error message

The largest charge group contains 47 atoms. The maximum is 32.

However, when using Gromacs2020, no error was displayed and the calculation was possible.
Can anyone please advise how to get this model working with Gromacs 4.5.7?

Best regards
T. Onot

Gromacs 2020 no longer uses charge groups due to a better nonbonded setup, so we just ignore it.

For version 4.1 you need charge groups, but such a large charge group is an astronomically bad idea, since it would mean atoms might be close to each other without interacting.

You can remove the charge grouping by assigning a separate charge group to each atom. That will work just fine if you use PME.

Cheers,

Erik

1 Like

Dear lindahl

Thank you for valuable advice!
And sorry for my delay in thanking you.

I understand that I have to assign a separate charge group to each atom.
Are you saying that I should rewrite [ atomtypes ] of itp file?
If so, do these values need to be redetermined in Gaussian etc.?

(I attached my calculation files.)
http://onot.sakura.ne.jp/test.zip

Sorry for my repeated questions.
I would appreciate it if you would give me some advice.

Best regards
T. Ono