The idea of computational alanine scanning is that a mutating single
residue from XYZ to ALA is a minimal perturbation that does not affect the
overall conformational ensemble. Therefore, a reasonable approximation is
to use the trajectory you generated in step 1 in place of the trajectory
you generated in step 2. You still need a topology file of the alanine
mutant, but once you have that, you can feed MMPBSA.py the mutated complex
and receptor files (or complex and ligand files if the mutation is in the
ligand), after which MMPBSA.py will automatically figure out which residue
has been mutated, then generate a matching trajectory file in which the
'new' alanine is in exactly the same conformation as the original residue.
This procedure eliminates the (long) MD step in your step #2. See the
MMPBSA.py paper for some more details on the implementation and method.
HTH,
Jason
On Mon, Dec 17, 2012 at 5:01 PM, Jonathan Gough
<jonathan.d.gough.gmail.com>wrote:
> Dear Guru's,
>
> I'm not sure I completely understand the alanine scanning
> experiment/algorithm in MMPBSA.py and if anyone can set me straight, that
> would be AWESOME.
>
> Here is what I THOUGHT i needed to do.
>
> 1. create a system of 2 proteins. save prmtop and inpcrd for unsolvated
> complex and individual proteins.
> - solvate, minimize, heat, density, equilibrate, and then do a production
> run to get snapshots. (MMPBSA calculations can be done on this alone
> obviously)
>
> 2. mutate 1 residue from one protein. save prmtop and inpcrd for
> unsolvated complex and individual proteins.
> - solvate, minimize, heat, density, equilibrate, and then do a production
> run to get snapshots. (MMPBSA calculations can be done on the mutant alone
> obviously)
>
> 3. Run MMPBSA using both sets of MD snapshots to find the difference
> between the 2.
>
>
> in re-reading I am getting the impression that I didn't really need to do
> step 2. Is that correct?
> Sorry if this is basic - I think i have just confused myself...
>
> Can I compare the energy between the 2 different MD runs directly? or
> would I be missing something if I did?
>
> Thanks,
> JOnathan
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Dec 17 2012 - 17:30:03 PST
