AMBER Archive (2005)

Subject: Re: AMBER: D-enantiomers

From: Thomas E. Cheatham, III (cheatham_at_chpc.utah.edu)
Date: Tue Sep 20 2005 - 21:36:34 CDT

> I am relatively new to Amber, but have not found any information on the
> following topic. Is there an easy way to implement the D amino acids in
> Amber? I am looking to simulate a short peptide which contains one or more
> D-enantiomers along the backbone.

As enantiomers have equivalent properties (outside of a chiral environment
other than bending plane polarized light), one would not expect that the
charges or intra-molecular parameters would be different for D vs. L in an
empirical force field. Moreover, there is no "restraint" or constraint
that enforces a particular chirality. The chirality is maintained by the
geometry. If you build the initial model in a D-amino acid conformation,
there is no (easy) way it will switch to L- during the course of the
simulation. [In fact, this used to be a problem with DNA adding H1' atoms
that were absent since at one point the structure in the library was
incorrect such that the hydrogen was in an initially bad/high energy
position where if you minimized this model-built structure, the H1' could
"flip" to the wrong stereochemistry completely messing up the structure].

The key is building the initial model; LEaP is set up to handle the
standard chirality (i.e. if atoms are missing, it will use its expected
geometry). If you provide the geometry yourself, everything should be
fine...

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