[gmx-users] RE: Protein is moving out of box

[Vitaly Chaban]

Well, the way you should follow depends much on what you want to
calculate exactly.

Of course, any artificial constraint is bad.

> It is even worst! Removing the center of mass motion of your protein
> alone as suggested is an aberration and should never be done!

> That doesn't necessarily "fix" the "problem".
> Which position of the
> protein is taken as a reference depends on the position of the first
> atom. Gromacs wants to have the first atom of a molecule in the
> rectangular unit cell at the origin. Even if you set comm removal for
> the group, the initial orientation and/or conformational changes may
> cause the first atom to jump over the boundaries.
> 
> Whatever the case, everyone doing MD simulations should be aware of
> periodic boundary conditions. There is no "outside of the box".


-- 
Vitaly V. Chaban
School of Chemistry
National University of Kharkiv
Svobody sq.,4, Kharkiv 61077, Ukraine
email: chaban at univer.kharkov.ua
skype: vvchaban