Hello Carl--
>
> I have been trying to get the script for docking 2 protein using
> chemical shifts and RDCs to work. One of my proteins contain Zn2+
> ions and I've been able to generate psf-files for this structure
> with some help from this mailing list and by modifying topology and
> parameter files. After running the dock_tor_rigid.py and comparing
> experimental RDCs to RDCs calculated from the complex, there is very
> little correlation. Is the alignment tensor allowed to vary in this
> protocol or is fixed from the start and throughout the protocol?
That script (eginput/dock_dipolar_chemshift/dock_tor_rigid.py) is
rather old at this point, but it should work ok. The Da and rhombicity
of the alignment tensor is fixed, but the orientation is varied.
>
> Both of the protein structures in the complex are NMR-structures,
> solved without RDC-refinement, where the correlation between my
> experimental RDCs and those calculated from the structures are
> weak. Based on chemical shifts we know that there is no major
> structural rearrangement upon complex formation, so experimental
> RDCs should agree with the uncomplexed structures.
What sort of R-factors do you get from the two structures with the
complex data? How do the standalone R-factors compare with those you
get from docking? If the stand-alone fits are too bad, then you really
will need to fix up the structures before docking.
>
> I tried refining my structures vs the experimental RDCs by using the
> refine.py script from the gb1_rdc example and removing NOE, dihedral
> and hydrogen-bond restraints.
> I don't know if this is the best procedure however, for getting
> better agreement with RDCs, since you remove restraints which keep
> the secondary structure together.
Right- you will need more restraints, even if they're artificial
restraints. Since these structures are NMR structures, you could use
the original NMR restraints as well. In short, this procedure will
result in poor structures.
>
> Is there a difference in how RDCs are used in the gb1_rdc/refine.py
> vs the dock_tor_rigid.py script? Also, what is the difference in
> how the parameters are setup? The syntax is different, and I can't
> readily discern what is different.
Right. Because the docking script is so old, there are many
differences, but it does utilize the RDCs in the same basic way. We
can rework the docking script into new syntax, but I suspect there are
other issues to deal with first.
>
> In the dock_tor_rigid.py script I initialize both structures using the
> following commands:
> command("param @TOPPAR:parallhdg_new.pro @param19.ion @TOPPAR:par_axis_3.pro
> end")
> command("structure @1a5ra_cns.psf @1tota_cns_zn_t2a.psf @TOPPAR:axis_500.psf
> end")
> command("coor @1tota_cns_zn_t2a.pdb")
> command("coor @1a5ra_cns.pdb")
> command("coor @TOPPAR:axis_xyzo_3.pdb")
> command("coor copy end")
This seems reasonable.
best regards--
Charles
_______________________________________________
Xplor-nih mailing list
Xplor-nih@cake.cit.nih.gov
http://cake.cit.nih.gov/mailman/listinfo/xplor-nih
