You may use the NAMD package for such investigation. It worked just fine for me. In your case You may try to use different names for your double ligands. Also it would be very useful to show here the log of minimization. This may help to understand your problem.

Thank you. I shall try that out

Ayesha,

NAMD is very much user friendly. Double ligand is not a problem. You must have topology and parameter for your ligands and must combine your protein and ligand coordinates into single pdb file with unique and different residue names for ligands. You should visit NAMD website and run some tutorials on your own. For performing protein-ligand simulation you should go through the free energy tutorials also. I'm attaching the link of NAMD tutorial webpage. It is also better if you learn VMD as well. It is easy.

http://www.ks.uiuc.edu/Training/Tutorials/

Since you have a few ligands, it is likely you will have to extend the binding site of the protein. This should be done carefully because protein is quite easy to spoil. I would do it by slowly drawing ligands using restrains. Do it better not by optimization, but dynamics at low temperature. For example suitable protocol is «abalone\Scripts\ standard protocols\optimization by dynamics (B).js» of Abalone package. I guess you can easily write similar for AMBER.

http://www.biomolecular-modeling.com/Abalone/index.html

Please provide detail methodology and research article on multi ligand simulation