Charmm FF Topology for substituted Heteroatom Residue

GROMACS version: 5.1.4
GROMACS modification: Yes

Dear Gromacs users,

I am working on the preparation of modified starch by substituting a cationic group on amylose and amylopectin chains. To generate the topology of modified chains by pdb2gmx command, I need to add the substituted cation residue type in .rtp file for which I need the charge on atoms of cation and the changed charge on atom of the amylose/amylopectin chains at the the site of substitution. Further, I will also need the parameters for ffbonded.itp file to add the bonded parameters in charmm FF. I already tried to use the Charmm Gui server for generating the topology by ligand modeler ensemble for a cation substituted amylose trimer to get the charges and the bonded parameters, but I feel that the charges provided by it are not 100% correct and also the atomtypes assigned by it are not unique to glycan residue. (To check it, I submitted the non-substituted amylose trimer to both glycan modeler ensemble and ligand modeler ensemble and found out that there is change in charge on glycan residue atoms for both the cases. It might mean that glycan modeler ensemble recognizes the glycan residue and assign charge to it uniquely compared to ligand modeler ensemble) I also tried the swissparam site but it gives no results. Can anyone suggest me another way to generate the topology for cation substituted chain or to get the charges and parameters for it so that I can modify the .rtp and .ffbonded files?
The other methodology that is in my mind is to take the charges and parameters for glycan residue of substituted chain from glycan modeler ensemble and the charges and parameters for substituted cation from ligand modeler ensemble. However, in this methodology also there is still one problem and that is of the charge and parameters of cation substitution site on glycan residue. Any suggestions on how to choose the charge and parameter for that substitution site?