Anyone who is interested in QM/MM modelling and is confronted with problems obtaining proper pdb files for AMBER or other codes, is asked to read further.
One big time-consuming problem of QM/MM calculations is the conversion of some custom built model into some format that is understood by MM codes like AMBER. The conversion of a simple .xyz file (xmol) to a pdb file with the proper description of every atom (not HETATM) is not trivial, since the connectivity has to be unambiguosly defined and every atom has to be correctly assigned to the residues.
I have written some piece of software that is capable of this transformation, checking all the bonding, assigning atom types, recognizing aminoacids and identifying backbone as well as side chains. If you are interested in such a conversion, please write an email to "CoolEsza AT gmx.de".
The following requirements have to be met.
a) only single chains can be processed
b) N-terminus has to be R-NH3 or R-NH2, C-terminus has to be COOH
c) only aminoacids in the file and no HETATMS.