modelling homodimers

Hi,
I was wondering if there is any way to make a model of a homodimeric protein using I-TASSER? I-TASSER only models a monomer when I submit the sequence even though the PDB templates it uses are homodimers as they should be (for eg see submission: http://zhanglab.ccmb.med.umich.edu/I-TASSER/output/S62211/ ). Ordinarily just making a copy of the monomer would be ok but the the active site of the protein I am trying to model is at the dimer interface. So I am concerned that the side chain conformations of the I-TASSER models in the active site may not be accurate and simply making another copy and aligning to the homodimeric template used from the PDB will result in clashes.
If I can trim the sequence such that the total number of residues in the final homodimer assembly is less than 1500 (currently it is more than 1500) is there any way to model the homodimer using I-TASSER?
Thank you,
Shiven