Arthur Wang <arthur at ipc.pku.edu.cn> wrote:
>It is well known that the electrostatic interactions is one of the most
>important features in ligand-receptor binding. To calculate it, partial
>charges of both the ligand and the receptor (usually a protein molecule)
>are needed to be integrated into the Coulomb formula. If the ligand is a
>small, common organic molecule, relatively reliable values could be
>obtained by using semi-empirical algorithm. But the charges on the protein
>remain as a stumbling block. At present, many people use residue charge
>templates, which are derived from the quantum mechanical results of three
>peptides. The drawback of doing so is obvious.
>So my question is, is there any better way to get the partial charges on
>the protein? Any experience? Any ideas? Pointers to any literature are
>also appreciated.
the way that i've seen used most often (and have started using myself, in
protein-protein docking problems) is one algorithm or another of assigning
charges to residues through pKa calculations. usually this will involve (in
its simplest form) placing formal charges on only the titrable atoms in the
protein (oxygen atoms on carboxyls, nitrogen atoms on arginines, lysines and
histidines, etc.), calculating a potential field over the whole atom,
converting this potential field to pKa difference, and iterating the thing
until you get constant pKa's over all residues after an iteration. you use the
pKa to find the charge on each atom.
my current favorite software of the moment, scott northrup's macrodox, uses the
algorithm of matthew and gurd (methods in enzymology, 130:413). there are also
articles from yang et. al. (proteins 15:252) and gilson (proteins 15:266...yes,
these articles are in the same issue, back to back!) describing different
viewpoints on the same subject.
hope this helps. be sure to e-mail me if you need any further info...
reply-to: dcp at hobbes.biosci.ohio-state.edu
--
chuck pearson - dopearso at magnus.acs.ohio-state.edu
osu biophysics program, daddy of amelia catherine pearson.
*THIS ACCOUNT TO DIE SOON - STAY TUNED FOR DETAILS*
