vincent Chaptal wrote:
we are refining a low resolution (3A) membrane protein using coot
0.6_pre2228.
In alpha helical regions, coot in doing a very good job but we have more
trouble in loops or kinks in helices.
Typically, in those regions, we just have density for the main chain
carbonyls and the CB of the side chains.
My question is how to use real space refinement in those regions?
Sphere refinement, backrub rotamers, Ramachandran restraints, maybe
hydrogens. In fact most of my talk at HEC-2009 from a few days ago :-)
very often, the side chain of large residue (Lys, Arg, Glu, Trp...) is
moving either in the main chain density, or in a neighboring density and
the refinement is impossible.
bleugh. Sphere refinement.
My way around is to mutate those large
residues to Ala
Bleugh! Sometimes it is good to chop off the residue atoms ("kill
sidechain") and then move the residue and do a fill partial. No
mutations or tedious clicking required.
and back to the original when i'm happy with the
refinement, but that doesn't seem optimum.
What i would like is to have coot only take care of the main chain,
trying to fit carbonyls and CB in the blobs of the main chain.
will *set-refinement-move-atoms-with-zero-occupancy 0* do the trick?
No.
is it possible to have a button saying "perform real space refinement
only on the main chain + CB" ? (say in the R/RC control box?)
This is on the list.
Paul.
