-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hello Jeff--
>
> We have used eginput/PSF_generation/addAtoms.py recently to add MTSL (a spin
> label) to proteins. We find that when addAtoms.py is used with Xplornih
> 2.15, the backbone of the protein is not altered and occasionally the spin
> label side chains have bad bond distances and/or angles. When the same
> addAtoms.py and protein is used with Xplornih 2.18 we get no bad spin label
> side chain bond distances or angles but the tertiary and secondary structure
> of the protein is highly altered. We have been able to correct the bad
> distances and bonds angles obtained with Xplornih 2.15 by using simulated
> annealing and energy minimization procedures very similar to those found in
> eginput/prot_prot. Our initial attempt to restore the secondary and tertiary
> structure to our spin labeled protein obtained with Xplornih 2.18 yielded a
> good backbone structure but some bad spin label side chain bond distances
> and angles. We used a reference structure with the correct backbone fold and
> a harmonic restraining potential (between the reference structure and the
> spin labeled molecule backbones) with a simulated annealing and energy
> minimization protocol similar to that mentioned above.
>
>
this script (addAtoms.py) should not alter any known atom positions. I
just checked and in it's original context it works correctly. I would be
very interested in seeing your modifications to the script which causes
the problems (I'm referring to problems in the 2.18 version).
It is possible to carefully include a call to fixupCovalentGeom after
addUnknownAtoms to clean things up- but carefully use the sel argument
so that input atom positions are not modified.
>
> My main question: Is there a script which will impose a different and
> defined backbone structure on a protein and maintain reasonable side chain
> structure?
I think you'll find the attached, modified script to be more robust.
best regards--
Charles
#
# mutate a single amino acid - to derivatized version.
# read in coordinates of the original protein and build in the positions of
# the new atoms.
#
#
xplor.requireVersion("2.12")
xplor.parseArguments() # check for typos on the command-line
seq="""
MET LEU LYS LEU LEU ASP LYS ASN GLY PHE VAL LEU LYS
LYS VAL GLU ILE TYR ARG SER ASN TYR LEU ALA ILE LEU
GLU LYS ARG THR ASN GLY ILE ARG ASN PHE GLU ILE ASN
ASN ASN GLY ASN MET ARG ILE PHE GLY TYR LYS MET MET
GLU HIS HIS ILE GLN LYS PHE THR ASP ILE GLY MET SER
CYSP LYS ILE ALA LYS ASN GLY ASN VAL TYR LEU ASP ILE
LYS ARG SER ALA GLU ASN ILE GLU ALA VAL ILE THR VAL
ALA SER GLU LEU
"""
simWorld.setRandomSeed(5521)
import protocol
protocol.initTopology(('protein','cysp.top'))
protocol.initParams(('protein','cysp.par'))
import psfGen
psfGen.seqToPSF(seq,seqType='prot',startResid=117)
from pdbTool import PDBTool
PDBTool('start.pdb').read()
notKnown=AtomSel("not known")
protocol.addUnknownAtoms(dyn_stepsize=0.01)
protocol.fixupCovalentGeom(sel=notKnown)
xplor.command("write psf output=new.psf end")
PDBTool("new.pdb").write()
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Processed by Mailcrypt 3.5.8+ <http://mailcrypt.sourceforge.net/>
iD8DBQFHFMYxPK2zrJwS/lYRAo2aAJ9IM2y7rh7eDKcYQm+hNB25o+Ct8ACcCRP5
ass/UNZ/HdM16MIRwjs2yOg=
=NQRh
-----END PGP SIGNATURE-----
_______________________________________________
Xplor-nih mailing list
[email protected]
http://dcb.cit.nih.gov/mailman/listinfo/xplor-nih
