On Sun, Apr 21, 2013 at 12:04 PM, Anselm Horn <
Anselm.Horn.biochem.uni-erlangen.de> wrote:
> Dear all,
>
> I have a pdb file (and top/crd) of a system already solvated in an
> truncated octahedral box. Now I apply a minor change in the system (pdb
> file) to obtain a second system.
> How do I proceed to setup a box from the modified pdb file with the same
> dimensions but without resolvating the system, i.e. without using the
> solvateoct command in leap?
>
> The leap manual describes the "set box" command: Giving one additional
> argument sets a cubic box, whereas giving a list of three numbers gives
> the lengths of the sides in A.
>
> Is the following approach sufficient where I obtain the length XXX from
> the known system (crd file)?
>
> # Leap script setting a known box dimension
> a = loadpdb "a_oct_inp.pdb"
> set a box XXX
> saveamberparm a a.top a.crd
> # End script
>
I'm not positive that any command except "solvateOct" will properly set
box information up correctly for a truncated octahedron (and there is no
setOct command, to my knowledge).
Therefore, you should check the topology file you get after doing this with
ParmEd. Particularly, look at the "printPointers" command, which you
should use to confirm that IFBOX is set to 2 (which is the value necessary
for correctly imaging particles with a truncated octahedron). If it is set
to 1, change it to 2.
You should then check that the angles in the topology file as well as the
inpcrd file (the latter is more important, though other programs may expect
the data in the prmtop to be as consistent as possible) are 109.4712190 and
not 90. Then, of course, run some short dynamics to make sure the box is
'ok', and make sure you run enough NPT to equilibrate the box.
HTH,
Jason
Any comment is welcome!
>
If anyone else has more experience with this type of thing, hopefully
you'll get an easier recipe...
But this comment still falls under the 'Any' umbrella :)
--
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 Sun Apr 21 2013 - 20:30:03 PDT
