My template is a homodimer and I want to model my target also as a hmodimer. The sequence of the chain A of homodimer in the PDB file is like: MKQA.............G. The sequence of the chain B of homodimer in the PDB file is like: MKQA.............GK. The sequence of protein in the NCBI protein database is like: MKQA.............GK. I have studied how to model a protein as a dimer using modeller (http://salilab.org/archives/modeller_usage/2010/msg00273.html). But if there is a slight difference in the two chains, like in this case the K residue is extra, then what should I do? (It would show a sequence difference if  I use chain copy the same alignment in both) Please suggest.