I want to model a structure based on another one. Both are trimers. My
sequence (corresponding to the "unknown" proteins) has additional
residues at the beginning of each chain, and I have to force them to
make an alpha-helix. The problem is that when I try to add the
restraints using as template the original chains-including PDB file and
separating chains in the .ali file, the residues numbering seems to
start from the residue 1 of the template, and I can't impose anything on
my sequence.
I'm not sure if I fully understand your problem, but Modeller always
numbers the *model* residues starting from 1 - the template numbering is
ignored completely. This is, of course, so that it IS possible to add
restraints on parts of the sequence which aren't in the template, as in
your case. I suggest that you first build a model with no additional
restraints, and then look at the output PDB (or if you are impatient,
look at the unoptimized model - the .ini file - instead). You should be
able to see the residue numbers and chain IDs of the model then, and use
these in order to add your restraints.
Modeller should have no difficulty in building multi-chain models. Just
add a / chain break character between each chain in the model sequence.