Hello Leonidas--
> So I am trying to use refine.py script to refine a crystal structure
> with RDCs from one medium. Given that I have no NOEs nor any other
> restrain, I only kept the rdc as restrains in the script input. The
> result is a completely unfolded protein structure with hugh energy
> coming mainly from torsion angle term.
> I tried to use the tutorial which uses the gb1 protein, again removing
> the NOE and dihedral restrains left with only RDCs. The result was the
> same.
> So my questions are: is it at all possible to refine a crystal
> structure only with RDCs? Are the NOEs absolutely necessary? What if I
> add "artificial" h-bonds restrains at the helices? Is there any term
> in the refine.py script which increase the penalty when the refined
> structured deviates a lot from the crystal structure?
>
If you take eginput/gb1_rdc/refine.py and change the refRMSD setup to
from posDiffPotTools import create_PosDiffPot
refRMSD = create_PosDiffPot("refRMSD","name CA or name C or name N",
pdbFile='g_xray.pdb',
cmpSel="not name H*")
refRMSD.setUpperBound(0.) # can increase this if the RDC fit isn't so good.
refRMSD.setScale(1000)
potList.append(refRMSD)
This will restrain the backbone atoms of your structure to the
reference (g_xray.pdb in this case), but should allow things to move
enough to satisfy the RDCs. An alternative is to generate distance
restraints from the reference structure, but this approach is somewhat
simpler.
best regards--
Charles
_______________________________________________
Xplor-nih mailing list
[email protected]
https://dcb.cit.nih.gov/mailman/listinfo/xplor-nih
