AMBER Archive (2009)

Subject: Re: [AMBER] AMBER job restart problem

From: Thomas Cheatham (
Date: Sat Aug 15 2009 - 23:44:27 CDT

> Thanks a lot! Dr. Cheatham. If I run constant (isotropic) pressure
> simulation, will the situation be better? My system is a lipid bilayer

The same.

> Another question is, if I don't use the original refc restraint
> coordinates all the time, but change it every time when restart the
> simulation, then the structure of the restraint part of the system might
> change too much if I restart the simulation too many times, right? Will

Yes, but the change is small. If you are restraining a single large
protein, what is happening as it runs is that the center of mass (assuming
the default of NSPCAL=1) of the protein is shifting. Using the previous
restart will have the shift, but the coordinates will not have changed so
much. If you are restraining multiple molecules, this is a larger issue
which can iterate offline.

> >> GRP1
> >> 100.0
> >> ATOM 1 13609

Again, drop the force constants to 5-15 kcal/mol. This will still highly
restrain atoms 1 13609 yet avoid the need for vlimit restrictions.


AMBER mailing list