[gmx-users] Switch box type

[Yun Shi]

Hi all,

I did conventional MD of a protein in solvent in a dodecahedron box
for 60 ns, and now I want to take the end structure (protein with a
shell of solvent in close proximity with the protein at 60ns) and put
it into a rectangle box for SMD simulations. I used g_select on the
original trajectory without any pbc manipulation.

But I was not able to convert the MD end structure into the right
representation by trjconv and editconf: although it is easy to make
the protein whole, the solvent shell seems to always sit partially to
the edge of pbc box. I first tried "Suggested trjconv workflow" from
gmx website, and then tried trjconv -pbc nojump/ atom/ mol, -tranlate
x y z, -center etc., and related flags in editconf.

Can someone suggest another workflow of doing this?

Much appreciated.

Yun