On 11/25/2009 02:59 AM, Francesco Musiani wrote: > I would like to calculate an all atoms (including hydrogens) model of a > metallo protein containing zinc. In particular, > I would like to bind Zn2+ ion to ND1 atom of an histidine residues. I > try to replace the residue "H" with "6" in the > alignment (I referred to restyp.lib library found in modlib directory) > but I obtain the following error in the log file. > > getpdf__392E> Your library requires all bins zero; nbinx: 5
Hmm - you have run into an interesting Modeller bug. Modeller is trying to build phi/psi restraints on this residue. It knows that your '6/HSE' residue has defined phi and psi dihedrals just like a regular 'H/HIS' residue, but then it falls over trying to look up the statistics needed to construct the restraints, since it only has statistics for the first 20 "standard" amino acid entries in modlib/restyp.lib. I will fix this for the next release, but in the meantime you have two possible workarounds:
1. Edit modlib/resdih.lib, replace the "22" line (just before the definition for ALA, around line 72) with "21", then search for the "HSE 6" line and delete that and the six lines that follow it. Modeller will then no longer know that HSE has dihedral angles, and won't try to calculate restraints on them.
2. Build your model with HIS rather than HSE residues, then manually move the proton in the generated model(s).
Ben Webb, Modeller Caretaker