For simulations at constant pressure, AMBER presently rescales the
positions of all atoms, regardless of whether or not they are
constrained (harmonic or belly restraints). This sometimes results in
some very strange behaviour. It would be desirable to be able to turn
this off.
Unfortunately, changing this is difficult because, presently, the
forces acting on these atoms are not ignored when calculating the
virial. If we allow the user to to freeze the immobilised atoms
(while rescaling the other atoms), then we must modify the way the
virial is calculated.
I have attached a document that suggests how to do that, along with
some code samples.
There are two equivalent ways to do this:
1) Include the restraint forces (the forces that hold the atoms in place) in
the virial. (This works for both rigid and belly constraints.)
2) Equivalently, one can simply ignore the forces that are exerted
on immobile atoms when calculating the virial.
(This works best for belly-constrained atoms.)
Immobilising restraints are most useful when applied to molecular
assemblies that would otherwise be unstable. Unfortunately, this is
exactly when AMBER pressure regulation fails. (See attached figure.)
I have included some code to override the default AMBER pressure
regulation behaviour in appendix A. From personal experience, I
realise that these kinds of posts can be very useful to people who
arrive at the AMBER mailing list archive from a google query. If the
attachment I sent does not make it to the archive, I'll repost the
code in the message body later on. (If the code is wrong, let me
know.)
Cheers
Andrew
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Mar 07 2011 - 05:00:05 PST
