[gmx-users] box vectores

[Nuno R. L. Ferreira]

Hello again.

Anton Fenstra wrote:
> I've seen one of the Gromacs tools (trjconv or editconf, IIRC) with an
> option to draw atoms at the corners and 'conect' statements in the .pdb
> file to represent the vertices. But I don't remember if that version
> made it into the distribution (it may have become a hidden option, check
> with 'trjconv -h -hidden' or 'editconf -h -hidden'), otherwise I'm sure
> Berk Hess could answer that...

 editconf has a -visbox flag, which (citing the man):
 visualize a grid of boxes, -1 visualize the 14 box images.
 I played around with this flag, but was not very successful.
-visbox -1 produces a file named images.pdb, and editconf gives a
segmentation fault. Anyway, this file has 15 lines , with no
'connect'entries.
-visbox 1(or 1 1 1) produces a pdb outputfile with 24 new atom lines and
also 'connect' entries for this 24 atoms. From the coordinates of these 24
new atoms, I can see they have resemblences with the box coordinates, but
after reading this pdb , the expected box shape is not obtained.
By the way, what are those values in -visbox flag? angles, vector
coordinates (which?) ...

> Another option is something like '-ur compact', which is an option to
> trjconv (which can also work on single .pdb or .gro files!), that will
> 'pack' your water molecules into the 'compact' representation of your
> box, corresponding to the intuitive rhombic box-shape you would want to
see...

Yes, I already knew this flags from previous e-mails on this list. I
used -ur compact -pbc whole to see the trajectory in VMD. And I was able to
see the rhombic dodecahedron box.

Perhaps I didn't explained myself clearly in the last e-mail. I want to
connect the vertices to see clearly the box shape. I was able to do this
with cubic boxes (it's easy with some tcl in the middle), but with such
triclinic boxes, I don't know how from 3 vectores only (page 13 of the
manual) I could draw all the vertices . I mean, which are the combinations
of vector coordinates to obtain the aimed polihedra? Where can I find this
kind of information? Perhaps in a polihedra book. Never had the chance to
study this kind of stuff.

Thanks Anton.
NUno