Hi there,
If there are a few clashes (acceptable - usually surface loops that
"enter" the surface part of another molecule or subunit) then you can
delete the parts of loops that usually have poor density and that clash,
refine, compute a new map and see if the loop can be traced (usually it
can be).
If there are really severe clashes, then your "solution" is not a proper
solution.
Also, have you checked the packing of the molecules in the crystal? Do
you have contacts (in the 3 directions of space) that explain the
crystal itself, with solvent regions or channels? This is also
indicative of a real solution or if your "solution" to the MR problem is
garbage.
Further, if you can have a search model (located somewhere in the PDB)
that is already a homodimer, then the molecular replacement calculations
should be repeated. "Better" results simply because there are more
vectors involved.
Fred.
Xianhui Wu wrote:
Dear All,
I am trying to resolve a protein structure with the use of
Molecular Replacement. However, some part of protein are overlap in
the interface of homodimer. Would someone please give me suggestions?
Thank you!
--
Best regards,
WuXH
