[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
PROTIN/REFMAC refinement with multiple conformations
*** For details on how to be removed from this list visit the ***
*** CCP4 home page http://www.dl.ac.uk/CCP/CCP4/main.html ***
Dear CCP4 folk,
Following up on Rui Zhao's question about refining with multiple
conformations:
I'm in the process of refining a structure in which several residues
appear to have multiple conformations. I have been able to get it
to work, but I have a question about what is the best way to set up
the input PDB coordinate files. (This is a long e-mail message, but
hopefully it is clear.)
Is it important for protin that the atom sets for the individual
conformers are in a block, as in:
ATOM 287 N LEU A 39 22.610 -15.414 0.006 1.00 29.06 7
ATOM 288 CA LEU A 39 24.048 -15.246 0.271 1.00 29.38 6
ATOM 289 CB LEU A 39 24.501 -16.202 1.371 1.00 32.32 6
ATOM 290 CG ALEU A 39 24.916 -15.618 2.725 0.50 34.69 6
ATOM 291 CD1ALEU A 39 24.369 -14.268 3.089 0.50 35.54 6
ATOM 292 CD2ALEU A 39 24.530 -16.645 3.794 0.50 36.39 6
ATOM 293 CG BLEU A 39 24.168 -15.856 2.826 0.50 34.69 6
ATOM 294 CD1BLEU A 39 24.672 -16.791 3.888 0.50 35.54 6
ATOM 295 CD2BLEU A 39 24.694 -14.441 3.082 0.50 36.39 6
ATOM 296 C LEU A 39 24.917 -15.333 -0.967 1.00 27.69 6
ATOM 297 O LEU A 39 26.036 -14.756 -0.937 1.00 29.26 8
or can they be interspersed as in:
ATOM 287 N LEU A 39 22.610 -15.414 0.006 1.00 29.06 7
ATOM 288 CA LEU A 39 24.048 -15.246 0.271 1.00 29.38 6
ATOM 289 CB LEU A 39 24.501 -16.202 1.371 1.00 32.32 6
ATOM 290 CG ALEU A 39 24.916 -15.618 2.725 0.50 34.69 6
ATOM 291 CG BLEU A 39 24.168 -15.856 2.826 0.50 34.69 6
ATOM 292 CD1ALEU A 39 24.369 -14.268 3.089 0.50 35.54 6
ATOM 293 CD1BLEU A 39 24.672 -16.791 3.888 0.50 35.54 6
ATOM 294 CD2ALEU A 39 24.530 -16.645 3.794 0.50 36.39 6
ATOM 295 CD2BLEU A 39 24.694 -14.441 3.082 0.50 36.39 6
ATOM 296 C LEU A 39 24.917 -15.333 -0.967 1.00 27.69 6
ATOM 297 O LEU A 39 26.036 -14.756 -0.937 1.00 29.26 8
I have actually tried both ways and the refinement did not give identical
results. The positions of the atoms from the two refinements were
extremely close but the R-factors, correlation coefficients and FOM's in
the refmac output were all worse at the beginning of each refmac cycle
when the atoms were interspersed than for the trial with the conformers
in a block. Conversely, the values at the end of the cycles were very
similar for both methods. Do the protin/refmac experts "out there"
have an explanation or other suggestions?
The structure I'm referring to is at 1.7A resolution with R=21% and Rfree
at 24%, 292 residues + about 100 waters, in which 16 residues appear to
have multiple conformations. The main reason I thought of testing both
methods is that it is much easier to set up the files with the atoms for
the two conformations interspersed. I have two nawk scripts, one to
generate two separate files for rebuilding in "O" and one to reconstruct
the single file for refinement. The latter works without further editing
if the atoms can be interspersed safely. In fact, it seems to work, but
I'm surprised that the two refinements were not identical.
These two scripts can be seen found at:
http://asterix.crchul.ulaval.ca/~rlc/work/scripts/
and others will be added when I get the time. These work for me, but if
anybody finds bugs, I'd be happy to hear about them.
A second, related question: while the electron density only shows
evidence for the multiple conformations of the side chain, and thus I
allowed only the atoms from the "gamma" position onward to exist as
multiple conformations, is it better to include more atoms in the two
conformations (in terms of getting better geometry perhaps)?
Cheers,
Robert
--
Robert Campbell, Ph.D. (418) 656-4141 x6348
Lab. d'Endocrinologie Moleculaire, FAX: 654-2761
Centre de recherche du CHUL, Univ. Laval, Sainte-Foy, Quebec, CANADA
<rlc@crchul.ulaval.ca> http://asterix.crchul.ulaval.ca/~rlc/