Hi All
I'm trying to include a cisProline in water refinement calculations - I'm using
a modified version of the script found in eginput/gb1_rdc/wrefine.py.
In earlier posts on this matter its suggested that there are two ways of doing
this:
1) load a pre-generated PSF using protocol.initStruct, and then load
the structure with protocol.initCoords instead of protocol.loadPDB.
I've used this method for the refine.py script loading in a single start
structure and it works.
If I try to do something similar for the wrefine.py script:
from psfGen import seqToPSF
protocol.initStruct("../generate/cov.psf")
protocol.initCoords(inputStructures[0]) - have tried with a single structure as
well - protocol.initCoords("../refine/refine_1.sa")
xplor.simulation.deleteAtoms("not known")
I get a lot of %CODANG-ERR: missing angle parameters errors and CODIMP-ERR:
missing improper parameters in the log file and the run aborts.
2) load the structure with protocol.loadPDB, and then correct the PSF
information using psfGen.cisPeptide.
In the script I have:
import psfGen
protocol.loadPDB(inputStructures[0],deleteUnknownAtoms=True)
psfGen.cisPeptide(110)
In the log file I get the following error message:
X-PLOR> patch CIPP
PATCH> reference=-=( resid 110 and segid "")
SELRPN: 19 atoms have been selected out of 2405
PATCH> reference=+=( resid 111 and segid "")
SELRPN: 14 atoms have been selected out of 2405
PATCH> end
%PATCH-ERR: atom C not found
and the Proline is then trans.
If I try the same thing in the refine.py script it works without error and I
get a cis Proline.
For the 2nd case if I move the import lines of the protein and apply the
psfGen.cisPeptide before the Nilges topology/parameters for water then the CIPP
patch is done correctly.
However this then causes a major problem in the actual refinement. Comparing
two structures calculated on a non-cis proline with just the position of when
the structure is imported (either before or after the water parameters) with
Molprobity I get the following:
For water params then loadPDB - ClashScore = 12.73, Poor Rotamers = 2,
Ramachandran Outliers = 1, Ramachandran favored = 91.75%
For loadPDB then water params - ClashScore = 34.55, Poor Rotamers = 9,
Ramachandran Outliers = 6, Ramachandran favored = 86.6%
What's also interesting is that the reported DIHE violations in the pdb
structure drops from 2355 (water, pdb) to 351 (pdb, water).
any help is greatly appreciated
Richard
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://dcb.cit.nih.gov/pipermail/xplor-nih/attachments/20130401/97e3bece/attachment.html
