Help finding proteins in the PDB...
- From: Ray Crawford <iguana "-at-" one.net>
- Subject: Help finding proteins in the PDB...
- Date: Tue, 22 Sep 1998 17:53:28 -0400
Howdy all!!!
I had a question about the contents of the PDB. I?m currently doing an
examination of a macromolecular docking program and I am looking for various
ligand/protein complex test cases. Ideally I would like to find two proteins
which have 5-10 experimentally determined structures, each with different
ligands
(ie Dihydrofolate Reductase with 5-10 different experimentally determined
ligands
in the active site). The two examples I was thinking about were:
1.) a protein with an active-site-surface cleft such as cyclophilin A.
2.) a protein with a partially or completely enclosed pocket like the retonioc
acid binding protein RXRg.
So far, I have looked at two examples -- ~10 experimentally determined HIV
protease complexes (which have not been deposited in the PDB) and 4 cyclophilin
A
complexes. The problem with the HIV protease case is that the active site is
too
darn symmetrical and I think that it is not a fair test of the program?s
?docking? ability ? there are too many ?good? orientations for the ligand (i.e.,
the program finds plausible binding modes but they are all WAY off from the
experimental binding mode). CypA, on the other hand, is a perfect example of a
cleft binding site but there are only 4 structures of small, drug-like molecules
experimentally determined in the PDB (and they are all dipeptides which, I might
add, the program nails right on?) ? I think that this example is too easy.
So, my question is two-fold. Do you all think I am neglecting any important
?binding modes? by limiting my examination to surface clefts and enclosed
pockets, and does any one know of any proteins in the PDB (or coordinate sets
you
would be willing to share?) that fit these criteria?
Thanks,
Ray Crawford
Research Associate
The Procter & Gamble Co.