Re: Converting a PSF file from one forcefield top another

From: Ilya Chorny (ichorny_at_gmail.com)
Date: Sat Oct 13 2007 - 15:01:42 CDT

I actually wanted to switch to OPLS. I was able to do it by partitioning my
system into the segments. But when I run the simulation I get an error
regarding the mixing rules. OPLS is geometric while charmm is arithmetic.
Is their a fix for using charmm for the lipids and opls for the protein and
water?

Thanks,

Ilya

On 10/13/07, Christopher Harrison <char_at_ks.uiuc.edu> wrote:
>
> oops . . . just a quick note . . . we're using the term "forcefield"
> here but we're really talking about which topology files are used
> during psf generation. The "forcefield" is composed of a combination
> of items including the topology, parameters (which are in the par*
> files called by NAMD's "parameters" command), and the functional form
> used (which is hard-coded into NAMD).
>
> Chris
>
> On Oct 13, 2007, at 2:43 PM, Christopher Harrison wrote:
>
> > Hi Illya,
> >
> > The easiest way may depend on how you generated the initial psf file:
> >
> > 1. If you used a psfgen script with the command line version of
> > psfgen then you can just change the topology files read at the
> > beginning of the script so they correspond to the forcefield you
> > desire and rerun psfgen. (I am assuming when you say different
> > forcefield you mean another version of the charmm forcefield.
> > Switching to amber, opls, etc forcefields is a different can of
> > worms.)
> >
> > 2. If you used the Auto-Psf-builder plugin in VMD, I believe you
> > should be able to just load your old psf and pdb then rerun Auto-
> > Psf. (In the Tkcon window: "mol new your-psf-file.psf" and "mol
> > addfile your-pdb-file.pdb".) Then start the Automatic Psf
> > generation by clicking Extensions -> Modeling -> Automatic PSF
> > Builder. When the Auto-Psf-builder window opens you'll need to
> > change from the default forcefield listed to a file(s) containing
> > the forcefield you wish to use. You can control segname
> > definitions in "Step 3: Chains" of the Auto-Psf-builder. Then
> > proceed as normal.
> >
> > Of course, there may be an easy way of which I'm unaware.
> >
> > Hope this helps get you started.
> >
> > Chris
> >
> >
> >
> > On Oct 13, 2007, at 1:14 PM, Ilya Chorny wrote:
> >
> >> I have a psf file which contains Lipid, water, and trimer protein.
> >> I want to convert it to a different forcefield. It there any
> >> simple way to do that? I tried separating monomers, lipid, water,
> >> and sodium but I get psfgen errors when I include the lipid,
> >> water, and sodium. My water and lipid are further broken down into
> >> multiple segments. Will I have to parse each segment?
> >>
> >> Thanks,
> >>
> >> Ilya
> >>
> >> --
> >> Ilya Chorny Ph.D.
>
>

-- 
Ilya Chorny Ph.D.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:45:23 CST