[gmx-users] truncated octahedron box vector angles and number of waters

[gromacs query]

Dear All

I am creating a truncated octahedron box and in vmd it appears as a
rectangular box. In mailing list it has been suggested to use trjconv -ur
compact option. But I have two issues,

1) I get box vector angle 70 90 70 with editconf and should be around 109
109 109 for a truncated octahedron box

editconf -f input.gro -o output.gro -c -d 1.2 -bt octahedron

    system size : 12.324 14.348  5.952 (nm)
    diameter    : 15.385               (nm)
    center      :  7.774  7.569  7.829 (nm)
    box vectors : 12.324 14.348  5.952 (nm)
    box angles  :  90.00  90.00  90.00 (degrees)
    box volume  :1052.41               (nm^3)
    shift       :  1.119  5.007 -0.568 (nm)
new center      :  8.893 12.576  7.261 (nm)
new box vectors : 17.785 17.785 17.785 (nm)
new box angles  :  70.53 109.47  70.53 (degrees)
new box volume  :4330.63               (nm^3)

2) I have a same system in AMBER and using truncated oct box with same 12
Angstrom water buffer but in Gromacs I get almost twice water molecules as
I get in AMBER but with AMBER I get 109 109 109 vector angles. Just to add
in AMBER I used the default value of closeness of solvent molecules being
1.0 Angs. I am not sure whats default in editconf (though I see it uses
default water density value)


Please suggest why I am getting different vector angles and number of
waters?

best regards
JIom