[gmx-users] PRODRG -cgnr

Atila Petrosian atila.petrosian at gmail.com
Sun Jul 27 18:20:21 CEST 2014

Dear gromacs users

I wan to use PRODRG server to do md simulation of my system (drug-protein

I know charge and cgnr parameters obtained from PRODRG (for ligand
molecule) are not accurate.

I know there are 2 ways to solve this problem:

1) If there are special functional groups in my ligand, I can find theses
functional groups being in protein, DNA and RNA molecules. Then I can
reassign charges and charge groups from these moieties.

2) Another way is to get partial charges from QM calculation.

I have some questions about second way:

Since the topology (itp file) parameters obtained from PRODRG are based on
GROMOS force field, I should consider cgnr correctly.

For example, in ffG43A1, N and H atoms form one charge group (cgnr = 0),
thus, summation of these charges are zero (-0.280 + 0.280 = 0).

N     N   -0.280     0
H     H    0.280     0

Using QM calulations, I can only obtain partial charges (mulliken charges).
My main question is that if I obtain partial charges using QM calulations,
how to set cgnrs?

More information about the gromacs.org_gmx-users mailing list