Re: [AMBER] Setting MM-PBSA script

From: Jag Silwal <jagsilwal.gmail.com>
Date: Tue, 25 Oct 2016 09:56:16 -0400

HI William,

thank you so much for your input.
Ideally what I would like to do is with the same MD run I would like to
calculate ddG for each of the ligands (as you have suggested) and also ddG
for dimer interface as well. Basically when I got that high ddG for my
first trial I attempted to use other protein as my ligand. But I ran into
few problems.
 When I prepared the individual PDB file, I simply edited in a text editor
and saved each as a separate PDB. For eg. in this case if I have four
chains, ABCD, then I made PDBs as: A, B, C, D,ABC, ABD, ACD and BCD. But
when I do this how would I handle the residue and atoms number issue? I got
error message right at the beginning of the calculation that the residues
are not sequential. How do I go about renumbering the residues?

Also, is that the correct way to split the PDBs? It would be great if you
could share your experience.

the last simulation I ran was with ABC as a ligand and D as receptor and at
least the MM-PBSA calculation was finished because residue number were
sequential in this case. I checked all the residues and atoms for each
individual PDBs and they all look ok to me. So I don't understand why the
ddG came up that high.

If you could share some details on how you handled your trimer I would
really appreciate your help.

Thanks again for your help

Sincerely,
Jag

On Tue, Oct 25, 2016 at 6:12 AM, William Lees <william.lees.org.uk> wrote:

> Just for info, I've been working with a trimer to which three ligands
> bind, and I can confirm that MM/PBSA and MM/GBSA analysis works as
> expected with this kind of setup. In the longer term, you'll probably
> want to run the calculations against both of the ligands so that you get
> a 'two for one' benefit from your trajectiories.
>
> Is it possible that you have defined the receptor/ligand boundaries
> slightly incorrectly, so that one or two ligand residues are in the
> receptor, or vice versa? I have found this easy to do and it does give
> rise to large delta delta Gs.
>
> All the best
>
> William
>
> On 24/10/2016 19:03, Jag Silwal wrote:
> > Dear all,
> >
> > I ran a simulation for protein complexes which includes four chains as
> > shown below. Protein "C" is a dimer and protein A binds to protein C and
> I
> > am interested to run MM-PBSA to estimate binding free energy between
> > protein A and protein C in its dimeric form.
> > Fo MM-PBSA, I created a separate PDB file for A and another PDB file with
> > the rest of the proteins (C-C dimer +another A). Then for the MMPBSA
> > command I treated A as the ligand protein and C-C dimer + A as receptor
> > protein. I ran with the following script:
> >
> >
> > &general
> > |startframe=4501, endframe=5550, interval=5,
> > |verbose=2, keep_files=2, strip_mask=':WAT,CL',
> > |/
> > |&gb
> > |igb=8, saltcon=0.150,
> > |/
> > |&pb
> > |inp=1, radiopt=0, istrng=0.15, fillratio=4.0
> > |/
> >
> >
> > When the run was done I checked the Delta delta G and it came to be about
> > 1000 kcal/mol which I think is way too high for the interface.
> > Previously, I have separately ran and analyzed the Delta delta G without
> > the dimer where there is only one A and one C the Delta delta G was -60
> > kcal/mol.
> >
> > So I am sure I am not either setting these different PDBs right or may
> be I
> > am missing something important while setting up for MM-PB/GBSA analysis.
> > How do I approach this problem?
> >
> > Any insight would be really appreciated.
> >
> >
> >
> > [image: Inline image 2]
> >
> >
> >
> >
> >
> > Sincerely,
> >
> > Jag
> > Graduate Student,
> > Michigan State University
> > Department of Chemistry
> >
> >
> >
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Oct 25 2016 - 07:00:02 PDT
Custom Search