On Mon, Feb 05, 2007, Mike Summers wrote:
> I'd like to calculated the chemical shift of a
> methyl group within a protein. Based on my reading
> of the Amber9 manual, I think I should use DivCon, but
> I can't find examples on how to execute DivCon,
> how to set up the input files, etc. Some examples I've
> found suggest that DivCon is executed from within Antechamber.
>
There are two ways in Amber that one can compute the chemicals shifts of the
protons in a methyl group. [Neither one is well documented :-( ]
1. You can use empirical formulas that compute ring current, electrostatic,
and peptide group contributions. This is basically the model in this paper:
%A K. Ösapay
%A D.A. Case
%T A new analysis of proton chemical shifts in proteins
%J J. Am. Chem. Soc.
%V 113
%P 9436-9444
%D 1991
In spite of its age, this actually does quite a good job for methyl protons in
proteins, (see Table III of the above paper) and (as far as I know) has not
been superceded by more recent work. I am pretty close to having a tutorial
prepared on how to do this in Amber. [The code in Amber is not broken, but
it's a bit tricky to prepare the needed input -- that's why my promise to get
this done is not quite there yet.]
If you just want a result for a few conformations, you can use the web
service at
http://www.scripps.edu/mb/case/qshifts/qshifts.htm, or download
the SHIFTS code (same web site) and run it locally.
2. You can also use the semiempirical quantum chemical code (DivCon) in Amber
to do this calculation. This would give estimates for both carbon and proton
shifts. The basic literature papers are here:
%A B. Wang
%A E.N. Brothers
%A A. van der Vaart
%A K.M. Merz
%T Fast semiempirical calculations for nuclear magnetic resonance chemical
%shifts: A divide-and-conquer approach
%J J. Chem. Phys.
%V 120
%P 11392-11400
%D 2004
%A B. Wang
%A K.M. Merz, Jr.
%T A fast QM/MM (quantum mechanical/molecular mechanical) approach to
calculate nuclear magnetic resonance chemical shifts for macromolecules
%J J. Chem. Theory Comput.
%V 2
%P 209-215
%D 2006
I haven't used this myself, but there is a sample calculation in
the amber9/test/crambin_qmmmnmr subdirectory. This may be enough to get you
started. Since this is a semi-empirical model, you'll want to read the above
papers carefully to get some idea of the expected accuracy.
...hope this helps....dac
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Wed Feb 07 2007 - 06:07:40 PST
