[gmx-users] about position restraints in optimize

[Erik Lindahl]

xu yechun wrote:

> Dear David,
>
> Maybe I did not express my question clearly. I use the same
> em.mdp(list below) and the same gro file but with the different
> content of posre.itp(first time: whole protein; second time: the
> mainchain of the protein; last time:only the Calpha of the protein).
> The rmsd ( last step structure to crystal structure) of every time is
> identical. So I think although I use "define = -DPOSRE" in em.mdp, the
> posre.itp is not be used in optimizing progress, otherwise I should
> get the different rmsd of every time. I guess that there is no
> position restraints in optimizing process whether using define =
> -DPOSRE or not.
>
> cpp = /lib/cpp
> define = -DPOSRE -DFLEX_SPC
> constraints = none
> integrator = steep
> nsteps = 100
>
Hi Xu,

Unless you have changed the topology manually, it shoud be -DPOSRES
(note the final S). This is just a flag that is added to the
preprocessor command, and in your topology there should be a directive like

#ifdef POSRES
#include "posre.itp"
#endif

that includes the actual position restraint file. Check the output of
grompp - it should tell you explicitly that it is using position
restraint coordinates from your input file.

Cheers,

Erik