Francesco Pietra wrote: > (1) What does "1SG" at columns 74-76 (PDB numbering) mean? > (2) Why serial numbers (Atom number + 1) are reported at column 80 > (PDB numbering)?

This is old-style PDB format, in which these columns were used to report the chain and atom serial number. New-style PDB format uses these columns to report the element of each atom, to resolve ambiguities in atom names (does HG1 mean mercury-1 or hydrogen-G1 ?). Modeller does not currently output new-style format for a variety of reasons (e.g. we don't know the elements for BLK ligands, standard CHARMM topology does not map atom names to elements, nobody has asked for it).

> (3) Should serial numbers said at (2) above be updated (like the Atom > number at column 11) when concatenating the model pdb files?

It's probably not necessary to update either column 11 or column 80. Most programs (including Modeller itself) ignore the serial numbers at column 80 when reading PDBs anyway.

> (4) Why the Chain identifier is not printed in model pdb files (which > is obviously needed for concatenation), or there was a way to specify > it be printed?

See my previous post. Actually, a chain ID is not necessary to concatenate PDBs - a TER is sufficient, if you are only using ATOM records.

> (5) When programming Python I take for obvious that it starts reading > the columns from zero, not 1. OK?

That's a very broad statement, but in general array indices start from 0 in Python, yes.

Ben Webb, Modeller Caretaker