homology model of protein dimer

Hi,

I am trying to generate a homology model of a protein dimer.

I was thinking that I could just generate the model of one subunit, and then superimpose it to both subunits of my template.

However, when I do this, there are a lot of steric clashes between my model subunits (about 19,000)...

I was wondering, is there maybe another way of generating a model of the dimer, so that clashes between the chains are avoided?

Thanks a lot in advance,

Cheers