AMBER Archive (2009)

Subject: Re: [AMBER] distance restraint and SHAKE

From: Thomas Cheatham III (tec3_at_utah.edu)
Date: Tue Jan 27 2009 - 12:16:30 CST

> I am running MD simulations of a patch of helices. I would like to keep
> these helices "helical" during some umbrella sampling calculations. I
> am currently using standard NMR restraints to hold the O-H atoms around
> 2Å. Although now that I think about this, it probably is not a good
> idea to use the SHAKE algorithm concurrently.

Others may chime in, but if the restraint forces are small (i.e. force
constants are in the 1-5 kcal/mol-AA range), this should be OK. You need
SHAKE on hydrogens for most explicit water molecules, the protein force
field, and likely any membrane environment you may be including. If you
are getting SHAKE failures during MD, reduce the force constant on the
restraints. Note also to maintain helical conformations, you could also
play with dihedral restraints on phi and psi (or for that matter use ff94
which is seriously helically biased better however would be ff99SB or
ff03 with weak restraints on the i->i+4 and maybe phi/psi angles).

Good luck. --tec3

_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber