Dear all,
I am currently helping with the implementation of the CHARMM forcefield
within AMBER along with Mike Crowley and Ross Walker. It is getting
there, and I am currently working on CMAP energies and gradients.
Here is a sample of what chamber (the tool for converting CHARMM psf and
parameter files to AMBER) is outputting at the moment to a CHARMM
enabled prmtop file:
%FLAG CHARMM_CMAP_COUNT
%COMMENT Number of CMAP terms, number of unique CMAP parameters
%FORMAT(2I8)
157 4
%FLAG CHARMM_CMAP_RESOLUTION
%COMMENT Number of steps along each phi/psi CMAP axis
%COMMENT for each CMAP parameter
%FORMAT(20I4)
24 24 24 24
%FLAG CHARMM_CMAP_PARAMETER_01
%COMMENT C NH1 CT1 C NH1 CT1 C NH1
%FORMAT(8(F9.6))
0.126790 0.768700 0.971260 1.250970 2.121010 2.695430 2.064440 1.764790
0.755870-0.713470 0.976130-2.475520-5.455650-5.096450-5.305850-3.975630
-3.088580-2.784200-2.677120-2.646060-2.335350-2.010440-1.608040-0.482250
...snip...
%FLAG CHARMM_CMAP_PARAMETER_02
%COMMENT C NH1 CT2 C NH1 CT2 C NH1
%FORMAT(8(F9.6))
-0.549160-0.535500-0.588110-0.754620-0.679290-0.038150 0.298460 0.326040
-0.375610-1.704360-3.061280-3.956460-3.576280-1.038930 2.012450-1.714610
-0.377660 0.317310 0.294580-0.042920-0.676620-0.744600-0.586590-0.554770
...snip.....
%FLAG CHARMM_CMAP_INDEX
%COMMENT Atom index i,j,k,l,m of the cross term
%COMMENT and then pointer to CHARMM_CMAP_PARAMETER_n
%FORMAT(6I8)
18 20 22 37 39 1
37 39 41 48 50 1
48 50 52 67 69 1
...snip....
Having looked through the nxtsec.f code, I have some general question
about comments in prmtop files:
a) Does anyone mind me being verbose in a following %COMMENT line(s) to
%FLAG as to what the %FLAG is describing?
b) Can a %COMMENT line follow a previous one? (I personally would like
to take around four lines to explain exactly what is going on and
perhaps even go some way to help to remove the need for the
http://ambermd.org/formats.html#topology file ;) )
c) Is having a specific flag for each CMAP term (%FLAG
CHARMM_CMAP_PARAMETER_n) a good strategy? Looking through the amoeba
code, that Tom Darden used an approach like this.
d) Any general pointers and/or hints+tips on the right thing to do here,
and what to avoid?
Thanks,
Mark
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Wed Feb 11 2009 - 01:13:03 PST
