Re: AMBER: Simulating a section of a molecule

From: Thomas Cheatham III <>
Date: Thu, 15 Feb 2007 14:33:40 -0700 (Mountain Standard Time)

> I am trying to make a 3D structure of an alpha helix, which is part of a
> My "alpha helix ", when simulated, doesn't look like an alpha helix at
> it's ends. The middle amino acids do resemble a helix though (it's 7
> a.a. total). I believe, in the true structure, the whole thing is a
> helix. Do you think this "improper" structure could be because I don't

Following up on Professor Simmerling's commentary, 7 amino acids is not a
very large helix. If I remember my literature correctly, at least 10
residues are necessary and possibly more (depending on sequence,
solubility, stability, solvent, etc.) for an isolated helix in aqueous
solution. Places to look would be papers by RL Baldwin, L Serrano, JM
Scholtz, S Marqusee, W DeGrado, etc.

Also, as pointed out by Simmerling, a given sequence may have a different
structure depending on context. The classic reference on this is Kabsch &
Sander PNAS 81, 1075-1078 (1984) showing pentapeptides have different
structures in different protein (contexts). This can be seen with 6-9
mers (Proteins, 30, 228-231 (1988); Proteins 41, 248-246) as well, and
even with an 11-mer (Minor/Kim, Nature 380, 730-734 (1986)). Further
complication is that the peptide may only become structured in a given
context (such as part of a protein or upon binding such as with some
transcription factors or hormones), i.e. it will be disordered in

So without further evidence (such as CD or NMR or detailed structure on
this 7-mer peptide) there is no way to prove or disprove that the MD
simulation results are "correct".

> Another question is, would you consider something to be a alpha
> helix/beta sheet, etc, even if programs (VMD, Chimera, DSSP, etc) don't
> call it one. Do you treat what the programs tell you as secondary
> structures as strictly correct?

This is a judgement call, akin in my mind to stating whether something is
hydrogen bonded or not... It depends entirely on your definition. To
understand this, you will need to delve into the literature and algorithms
used by the various programs to assign secondary structure. For some a
helix may simply mean sequential i->i+4 hydrogen bonds, to others it may
mean populating certain regions of phi/psi space, and to others it may
mean a certain CD signature...

There is a vast literature on peptide/protein folding and/or helical
propensity from simulation; look for work by Simmerling, Duan, Pande,
Daggett, Hermans, CL Brooks, J Shea, Hummer, Garcia, among others...

\-/ 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
/-\ (801) 587-9652; FAX: (801) 585-9119
\-/ BPRP295A
The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" to
Received on Sun Feb 18 2007 - 06:07:24 PST
Custom Search