[gmx-developers] Heads up! trjconv broken

[Tsjerk Wassenaar]

Hi,

> Do the trajectory formats allow the storage of rotated boxes? In that case
> any operation that rotates the coordinates should also rotate the box. (But
> then the PBC routines should also work with arbitrary rotated boxes!)

Of course. The box is saved as a 3x3 matrix, which allows specifying
any lattice in 3D. The only thing is that quite a number of routines
need to be updated. E.g., the main fitting routine (do_fit) needs to
get the box matrix as argument, and all calls to it need to pass the
box (or a dummy box).
But doing so makes the resulting boxes non-compliant with Gromacs
rules, making calls to PBC routines break. That means that all PBC
operations need to be modified too. One way would be to pretransform
box and coordinates back to easy to handle orientation prior to PBC
operations, and posttransform back to the original orientation.
Finally, grompp/mdrun should probably be made capable to rotate and
lattice-reduce a box to have the internal representation during the MD
run as optimized as possible).

Cheers,

Tsjerk



-- 
Tsjerk A. Wassenaar, Ph.D.

post-doctoral researcher
Molecular Dynamics Group
Groningen Institute for Biomolecular Research and Biotechnology
University of Groningen
The Netherlands