Regarding COACH server binding site prediction on Trimeric protein

Dear team,

JOBID: CH003952

I would like to use COACH server for predicting ligand binding sites in my heterotrimeric homology model with A, B and C chains. I found that submitting my homology model with default options to COACH server has considered only single chain (first chain A). If I need to make all chains to be considered during the run, is there any specific option available? Finding sites at interfaces of the monomers is my interest. So, please let me know how can I do with my protein to get the required job done.

Thank you

Regards,
Lalith