From: Francesco Pietra (
Date: Sat Mar 05 2011 - 14:38:45 CST

I was gathering information on H-bond in NAMD and read your notes of
last year about.

My specific interest is in how H-bonds are treated with colvars
component "hBond".

Naive question: in setting such an option for GLUP-GLU and
GLUP-chloride_anion hydrogen bonds, should the acidic proton of
neutral GLU be present in the PDB file and be specified with GLUPP
patch? Or nothing more is required than the distances between the two
oxygen atoms (or GLU oxygen atom and CLA anion) while the system sets
in between a dummy atom? If I understand, the FF does not care about
linearity (strong H-bond) or bending (weak H-bond).

Doubly naive question (as I come from another mm package): in setting
the GLUP patch, should the residue name (in the PDB file for autopsf)
still be called GLU, or a non-standard four-letter residue name GLUP
be used? I am asking that because the psf/pdb files I have generated
can be inserted into a lipidici membrane and NAMD minimized and heated
correctly but the ensemble is not as I wanted about H-bonds. So, I am
starting again from my X-ray diffraction data. Is any other method to
check the outcome of patching than graphically from the structure
besides examination of the PDB file?

Thanks a lot for advice

francesco pietra