AMBER Archive (2008)

Subject: Re: AMBER: improper dihedral in ff03 united atom forcefield

From: Bill Ross (ross_at_cgl.ucsf.edu)
Date: Tue Jul 22 2008 - 23:28:07 CDT


> will somebody give me some explanation?

Note that nonplanar impropers cannot be used with the Leap programs.
As posted earlier today:

Subject: Re: AMBER: amber ff and how it deals with out-plane chiral C
From: Bill Ross (ross_at_cgl.ucsf.edu)
Date: Tue Jul 22 2008 - 16:44:36 CDT

Before Leap was developed, improper dihedrals were specified by
listing atoms in the residue definitions. This allowed the order
of atoms to be uniquely specified in each case, so that for the
non-planar impropers in the united atom force field, the direction
of the offset from the plane could be specified.

With Leap, the decision was made to have impropers be applied just
like bonds, angles and torsions: by finding topologically appropriate
groups of atoms and looking for parameters for them. (The main
difference is that improper parameters for each group are not required.)
This eliminated the step of setting the sense by specifying the order
of the atoms. (It also implies that impropers will be applied to all
matching sites.)

Thus if a non-planar improper is specified, the order of the atoms
surrounding the central one will be chosen on the basis of alphabetical
order (I forget if by atom name or type), not chemical context. Finding
a solution for the united atom force fields was postponed indefinitely
since these weren't used much. I was the one who discovered and debugged
the problem, while preparing the original version of leap for release.

Bill

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
      to majordomo_at_scripps.edu