AMBER Archive (2006)

Subject: Re: AMBER: Simulating a periodic crystal

From: Thomas Cheatham (cheatham_at_chpc.utah.edu)
Date: Wed Mar 29 2006 - 09:57:10 CST


> I did some search in the Amber website and found some information with
> regard to this issue (in the attached files) Dr. David A. Case said
> "only non-covalent interactions between unit cells are allowed", to my
> understanding, this implies that Amber will not work for my case.
> However, from Dr. Thomas E. Cheatham, III's comments and several of Dr.
> T. A. Darden's papers [(Biophys. J. 78, 668 (2000), Biochemistry, 32,
> 1443(1993)], it seems Amber can do simulation in a periodic crystalline
> environment even part of a biomolecule crosses the unit cell boundary.

Both posts are "correct", although DAC's is more explicit with respect to
the treatment of infinite periodicity (of periodic helices like DNA or
actin). AMBER is not currently set up to handle infinite periodicity
across the periodic boundaries since there is no code to handle the
intramolecular interactions among the periodic images. Additionally,
AMBER only supports minimum image conditions within a "full" unit cell; in
other words, you cannot have only part of the periodic cell present and
impose periodicity on the rest (for example putting one half of a dimer in
the unit cell and autogenerating the symmetric second half) and you also
cannot have a direct space cutoff larger than 1/2 the shortest periodic
box dimension.

In the papers of Darden, a unit cell was replicated to mimic the crystal.
Either one or four (or eight?) copies of the crystal periodic unit cell
were explicitly present representing the crystal structure. This works
because there are no bonds between elements of the primary unit cell and
its periodic images (i.e. the system was not infinite).

Mimimally to get the infinite helix to work, code would need to be
modified to include the extra bonds between the periodic unit cells; this
would require either adding periodic imaging to the bond, angle and
dihedral routines or (as done in CHARMM) creation of virtual copies of the
images. Extending the code in this manner will not be trivial and will
involve changes to the prmtop (LEaP), sander (and accessory programs like
anal, nmode, ...).

In the earlier post, DAC mentioned that he believed that CHARMM can do
this; it can, at least within the CRYSTAL/IMAGE facility and it can also
impose symmetry, for example treating only 1/60th of a 60-fold symmetric
virus particle and imposing symmetry on the rest (see work by C. Post
and/or BR Brooks).

\-/ Thomas E. Cheatham, III (Assistant Professor) College of Pharmacy
-/- Departments of Med. Chem. and of Pharmaceutics and Pharm. Chem.
/-\ Adjunct Asst Prof of Bioeng.; Center for High Performance Computing
\-/ University of Utah, 30 S 2000 E, SH 201, Salt Lake City, UT 84112
-/-
/-\ tec3_at_utah.edu (801) 587-9652; FAX: (801) 585-9119
\-/ BPRP295A http://www.chpc.utah.edu/~cheatham

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu