MGiò wrote: > I'm trying and model a 7-helix trans-membrane receptor, as I don't like > the way MODELLER models my helices, while I highly appreciate its loop > modelling, I've been using the "model-myini.py" input script to model my > receptor starting from a initial.pdb containing the 7 helices already > modelled. Every MODELLER run ends with the "nselat__596E> Atom index is > out of range" error. Considering that this error changes the couple of > atoms referring to any time I delete the corresponding residue in order > to by-pass the error. I've attached to the mail the input files I used > and the log file. Can you help me?

OK, looking at the full set of input files, the problem is clear. The initial model you provide must match the sequence you're trying to build, but in your case the sequence has 2871 atoms, while the 'helices.pdb' file you're providing as an initial model has only 1702. I suggest you run your script again, but with the "inifile='helices.pdb'" option removed, so that you get ordinary model building, and look at the generated GHSR.ini initial model. Your helices.pdb file must have the same number of atoms (in the same order).

I'll add some extra checks to the code in future so that you get a more explanatory message than "atom index out of range".

Ben Webb, Modeller Caretaker