[gmx-users] Refcoord_scaling with Position Restraints on only a part of the protein

[Machtens, Jan-Philipp]

Hi,
I'm wondering how to optimally equilibrate
a system of a protein in aqueous or lipid environment.
Usually I use multiple steps for equilibration (using weak coupling schemes for P and T) and
(1) I first apply position restraints on all heavy atoms of the protein
and after "proper equilibration" ...
(2) I continue with an additional equilibration step with backbone-only posre on the protein.

Usually, for step (1) I use "refcoord_scaling=COM" to keep the internal geometry of the (reference coordinates of the ) protein unaffected by the barostat.
With this refcoord setting for step (2) the barostat will of course scale again the coordinates of all protein atoms, but the internal distances between the reference positions of backbone atoms remain constant. This appears a bit strange to me, since it could result in additional strain on the side chains and could actually "worsen" the protein structure. Would you therefore prefer "refcoord_scaling=all" for this step (2) with posre on the backbones only ?

Of course, this issue is not necessarily "relevant" since fluctuation in box geometry in step (2)
could be relatively low, if step (1 ) was sufficiently long.
Thanks for your opinion!
Cheers,
Jan-Philipp



------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------
Forschungszentrum Juelich GmbH
52425 Juelich
Sitz der Gesellschaft: Juelich
Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498
Vorsitzender des Aufsichtsrats: MinDir Dr. Karl Eugen Huthmacher
Geschaeftsfuehrung: Prof. Dr.-Ing. Wolfgang Marquardt (Vorsitzender),
Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt,
Prof. Dr. Sebastian M. Schmidt
------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------