Ah! Stupid Gromacs, why won't you centre my protein?
Molecules can be centred into a box using the trjconv_d command with gromacs
trjconv_d -f protein.trr -o protein-center.trr -s protein.tpr -center
-pbc mol -ur compact
and select centre on protein when prompted.
However, this doesn't work with a protein dimer as it centres the protein
based on the centre of mass, which is the same point if the dimer is together
or wrapped over the periodic boundary conditions, i.e.
is the same to Gromacs. This is of course technically correct with a periodic
boundary, but is very annoying for visualisation.
Therefore a slightly more complex method needs to be used. Using make_ndx_d a
separate indexed list can be created so that when prompted using trjconv_d we
can use our own one.
make_ndx_d -f protein.gro -o protein.ndx
The terminal will then become interactive with instructions. The best option
is to select one on the protein monomers by using atom numbers by typing
(assuming 3000 atoms in the monomer)
a 1-3000
q
and hitting enter after each line.
Next we can use trjconv_d with the .ndx file
trjconv_d -f protein.trr -o protein-centre.trr -s protein.tpr -pbc mol
-center -ur compact -n protein.ndx
and select the a 1-3000 option when prompted for centering (probably at the bottom)
and the system for writing out.
It may also be beneficial to remove any rotational and translational movement from
the protein if you are going to visualise it
trjconv_d -f protein-centre.trr -o protein-fix.trr -s protein.tpr -ur compact
-fit rot+trans
and select protein option for the R2 fit and system for writing out.
