On Mon, Mar 28, 2016, Dave Cerutti wrote:
>
> I've come to realize that, in GAFF, Junmei is playing a very sophisticated
> game. The parameters listed in gaff.dat aren't precisely what appears in
> the resulting prmtops. There are, instead, a rich set of rules listed
> elsewhere on our website by which those parameter prototypes get
> post-translational modifications to make a functioning system. In light of
> this, I'm trying to figure out how one can go about parameterizing a ligand
> with mostly GAFF yet introduce some specific modifications of his own (the
> new modifications would not be meant to go through the mixing and
> modification that other GAFF parameters do).
>
> Looking at the sustiva ligand parameterization tutorial and the general
> strategy with paramfit, it seems the things to do is write a frcmod file
> that then gets read in to supplement or overwrite the original parameters.
>
> So... is the answer that GAFF can be modified with a frcmod file, and tleap
> knows to apply various mixing rules and contexts to modify GAFF's prototype
> parameters, but if it then sees a frcmod file with relevant parameters it
> will apply those at face value, overwriting the GAFF-generated numbers?
I'm not sure I understand your question. The parmchk2 program uses the rules
you mention above to create a frcmod file specific to the molecule it is
dealing with. tleap knows nothing about "prototype parameters" or
"mixing rules and contexts". All tleap sees are these two statements:
source leaprc.gaff <-- reads in general parameters
loadAmberParams frcmod.<yourmolecule> <-- file created by parmchk2
tleap doesn't have any idea where the parameters came from, or whether they
are GAFF parameters or not: all those "smarts" are in parmchk2.
So, just load yet another frcmod file (after the two above) if you want
to make further modifications to the parameters that actually get into the
prmtop file. All parameters read into tleap are treated at "face value".
Later frcmod's over-write earlier ones.
Hope this helps...apologies if I have misunderstood your question and are just
telling you things you already know....
....dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Mar 28 2016 - 18:30:03 PDT
