psfgen: pdb and nucleic acid

From: David E. Konerding (
Date: Wed Dec 03 2003 - 13:52:52 CST


I am attempting to make a psf containing a protein bound to a DNA duplex
so I can simulate it in NAMD (the model is derived from 1A31.pdb and

unfortunately, the manual doesn't give much guidance in building
protein/dna PSF, here's what I did: first I split the PDB chains into
distinct files, and
massaged atom names and residue types to make psfgen happy with them. I
can write .psfgen scripts which work to make each chain into a distinct PSF
but I can't get the whole one, because, apparently, psfgen wants to
apply the default NTER and CTER patches to the DNA terminii.

Here's a couple examples of working individual psfgen input:

topology top_all27_prot_na.inp
alias residue HIS HSD
alias atom ILE CD1 CD
alias atom PHE OXT OT1
segment topoI { pdb topoI.pdb }
coordpdb topoI.pdb topoI
writepdb topoI.pdb
writepsf topoI.psf

##DNA chain C
topology top_all27_na.inp
alias residue A ADE
alias residue G GUA
alias residue C CYT
alias residue T THY
segment dnaC { pdb dnaC.pdb }
coordpdb dnaC.pdb dnaC
writepdb dnaC.psf.pdb
writepsf dnaC.psf

Note: I had to use the top_all27_na.inp to work with DNA, otherwise, it
applies the default amino acid terminal patch.

When I combine these into a single script, I have to select one topology
file or another; I picked top_all27_prot_na, since that seems
to have both proteins and nucleic acids, which would be necessary for my
mixed system.
Byut that's the input file that causes the DNA terminii to be patched.

I looked pretty extensively in the NAMD manual but it wasn't illuminative.


This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:37:12 CST