Re: Changing the Secondary Structure of a Protein

From: Peter Freddolino (
Date: Sun Dec 28 2008 - 08:48:38 CST

Hi Mike,

why do you want to save your changes to a file? You say "prior to
running a simulation in NAMD", but namd has no concept of secondary
structure, so even if those residues are initially flagged as helical it
won't affect your simulation (aside from starting from a helical

STRIDE (and thus VMD) has a purely structural definition of secondary
structure, so changing the HELIX string of the pdb (for example) won't
do anything; you need to change the structure if you want it to be read
in as something different. You could, for example, alter the backbone
dihedrals of the segment you want to make non-helical, or melt it at
high temperature with the rest of the protein restrained. Pretty much
any set of angles that breaks the backbone hydrogen bonding pattern
should work for your purposes. You would, of course, need to keep in
mind that in the absence of more information you'd be introducing an
essentially random structure in that region, with the only thing you
know about it being that it is (properly) not helical.


MIke S wrote:
> Hello,
> I would like to modify the secondary structure of my protein prior to
> running a simulation in NAMD.
> Currently when I load the (crystallographic) PDB in VMD certain
> residues (78-81) are helical. However, I would like to change these
> residues to be turns or coils since NMR data show that these
> residues are not in fact helical (ie. the conditions used in the
> crystallographic study promoted helical formation in this region of
> the protein).
> I've glanced at some of the previous messages posted on VMD with
> regards to STRIDE and making changes to secondary structure.
> I've tried the suggestion in this post
> which
> worked, however when I tried to save the corresponding PDB file the
> changes to the secondary structure weren't preserved.
> Obviously, using the 'structure' command only changes the graphic
> image of the protein in the VMD display...and my knowledge of STRIDE
> is slim. I was hoping to save a PDB file with the corresponding
> structural changes and follow that up with a run of psfgen to generate
> the necessary PSF and PDB files for my NAMD simulations.
> How can I make the necessary changes to the protein structure and save
> this to a file? I assume I will need to manipulate some of
> the data listed in the PDB or PSF files. In fact, I've played around
> with the HELIX string in the PDB file and modifyed the residues listed
> in the HELIX string of interest, but it didn't work - at least when I
> checked the structure on VMD the residues were still helical.
> Also, if I need to manipulate the dihedral angles for
> example, how do I know which phi and psi values to use for a coil
> or turn?
> Some side notes:
> There aren't any NMR PDB files of the full protein that I'm aware of.
> In addition to the above residues there are other residues that would
> need structural changes. I've used the program SwissPDB Viewer to add
> missing residues (1-4) and for some reason these residues are helical
> even though they should be coiled. (For those familiar with SwissPDB:
> I've tried to change the secondary structure using the control panel
> in SwissPDB before importing to VMD, but the structure remains
> helical). Btw, I'm using the latest versions of VMD and NAMD.
> I hope someone can offer some assistance.
> Happy Holidays,
> Mike
> (PS. I also posted this on the NAMD mailing list)

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:48:53 CST