[modeller_usage] regarding N terminus region

I have a protein sequence whose N terminus region does not matches with the template. Is it neccessary that N and C terminus regions of a protein should be fixed for homology modeling? The n terminus region contains 12 residues.  There are long gaps also within the alignment. Can these regions be modelled by loop modelling? I am also attaching the alignment file.