AMBER Archive (2002)

Subject: Fw: Do I have to rerun 1ns MD??

From: Lepsa (lepsik_at_marilyn.uochb.cas.cz)
Date: Mon Sep 23 2002 - 10:50:26 CDT

Dear Fan,
I've asked both these questions and here are the answers from Dr. Holger
Gohlke:

Hope this helps.

Martin Lepsik

> > I'd like to ask you how to proceed. It seems mm_pbsa doesn't like gaff
> > (lower-case) atom types (messages like bad atom type c3). At which stage
> > shall I make them parm99 (upper-case)? Or are there other solutions?
> It's only indirectly related to mm_pbsa - the setup routine for the SAS
> algorithm in sander was buggy. We fixed this - have a look at:
>
> http://www.amber.ucsf.edu/amber/bugfixes/7.0/bugfix.7
>
> Download the patch to amber7/src/sander, do "patch < bugfix.7" and
> re-compile / re-install sander. Then everything should work.
>
////////////////////////////////>

> try "199-199" as input for your decomposition. Rerunning the MD job
> should not be necessary then.

> I'd like to ask you for advice or opinion.
>
> I've run a 1ns MD in explicit solvent of a protein-ligand complex to
> describe the energetics of binding. Then, AMBER 7 MM-PBSA came to play
> and showed the possibility of decomposing the binding energy by residue.
> I'd be very happy to use it but as I've used a single-residue-ligand
> topology for my MD, I gather that I cannot use MM-PBSA decomposition
> (error message - Wrong format for decomp: 199) {199 is the ligand
> residue number}.
>
> Is there a way how to avoid rerunning the 3-week MD job and get the
> decomposed energies?