Protein design: Ligand binding between two units of a homodimer

Hello, 

I am trying to find mutations for a protein to bind a non-native ligand. The protein is a homodimer with ligand binding in between two monomers of the homodimer unit. The cyrstal structure of the protein bound to a native ligand is known. I am following the protcol from Moretti, R., 2016. I am not sure how to constrain the system so that I get the same set of mutations in both chains of the dimer unit. Are there any options I can use to add this constraint?

If I do not add any options or symmetry, I get a different set of mutations on both chains. I have tried making a symmetry definition file but I am not sure if it makes sense to use symmetry in this case.  I am using Autodock to dock the non native ligand in between the monomers and then delete one chain to make the Lig_positioned.pdb file. The single chain unit is used for design with the symmetry file. I get many errors in this case, possibly because the system is not inherently symmetric. 

Any suggestions or guidance on what would be a good method to solve this issue would be greatly appreciated. Thank you.