How to model the missing N-terminal residues (Residue number: 1-16) from
the protein structure?
This is straightforward; just add an insertion at the start of your
alignment (i.e. add 16 residues to the model sequence, and 16 gaps to
the template):
Note though that since Modeller has no template for this sequence, and
only one end is constrained by known structure, most likely you'll get
an extended chain. Since such a model is of little further use, it's
usually better just to exclude these termini.