On Wed, Nov 9, 2011 at 4:30 PM, David A Case <case.biomaps.rutgers.edu>wrote:
> On Wed, Nov 09, 2011, Jan-Philip Gehrcke wrote:
>
> > On 11/09/2011 04:22 PM, David A Case wrote:
> > >> p = loadpdb protein.pdb
> > >> l = loadpdb ligand.pdb
> > >> c = combine {p l}
> > >> [addions, solvateoct, saveamberparams]
> > >
> > > I agree that this ought to work as well, although it is far less
> > > common as an approach. Are the coordinates of p and l in the output
> > > incprd file the same ones as in protein.pdb and ligand.pdb? If so,
> > > this indeed sounds like a bug; if you can post a (small) example
> > > showing the problem, we can take a look at it.
> >
> > David,
> >
> > I can't exactly follow you here, because for me the "combine approach"
> > worked as expected -- without any bugs.
>
> OK...this is probably my fault, although both you and Senthil Natesan
> are writing on the same thread, and maybe I got things mixed up. But I
> suspect that I am not the only person that is now confused.
>
The way I understood it, Senthil described his problem, and his use of
"combine" was suggested as the culprit since it's unusual. Jan-Philip
merely jumped in to say "well, combine works for me, so I don't think
there's any overtly obvious bug in combine". But perhaps I'm mistaken, and
now I'm confused :).
FWIW, I also use the combine approach for protein-ligand systems (take a
PDB of the complex, split the ligand portion into another file, and use
"combine", which should preserve coordinates). This results in, as
Jan-Philip observed, "expected" behavior.
My analysis is that "combine" works perfectly for this kind of approach
(i.e. when you literally split apart a non-covalently-bound structure and
wish to use "combine" to put them back together) since it conserves
coordinates of each part (and maybe transforms them all at the end, I
haven't checked this). However, if you don't have several structures that
you pulled apart this way, then "combine" can give a whole host of issues.
For instance, if you wish to create a dimer from a monomer (or something
equivalent), then you can't do something like
mon = loadpdb monomer.pdb
dim = combine {mon mon}
While the prmtop you'll get is valid, the dimer will simply be 2
superimposed monomers (since the coordinates are the same). When you go to
"solvateoct", it will treat the system as much smaller than it really is,
since the coordinates of each atom only suggest the system is as big as the
monomer.
I've used the "combine" command several times and always got my expected
behavior.
All the best,
Jason
--
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 Wed Nov 09 2011 - 14:00:04 PST
