See also the PDB summary.

From a very simple autoPROC run of this Se-MET crystal, we have data for 2 wavelengths (infl and hrem). We also have the sequence (181 residues, 4 Met - 3 of which are N-terminal around a His-tag):


Initial autoSHARP run

We start autoSHARP through the default (httpd-based) interface and define the following for a 2-wavelength MAD experiment:

  • the sequence
  • explicitely setting spacegroup to P41212 (since this seems the most likely spacegroup - see autoPROC tutorial)
  • searching for 2 Se sites (seems a good compromise given the N-terminal Met residues in the sequence)
  • infl data:
    • wavelength 0.9794 A
    • f' and f'' values (guessed): -6 and +4
    • reflection file
  • hrem data:
    • wavelength 0.9184 A
    • reflection file

After 2 1/4 hours we have

10 chains with 439 residues : 5 docked in sequence (333 residues) [R/Rfree=0.2218/0.298]



Can we do better?

We could also try a new tool we're developing (combining density modification using PARROT and model building using BUCCANEER, both programs by K. Cowtan):

% ljs -mtz eden_flat_55.5pc.mtz -fast -seq ../../datafiles/1Y13.pir

which after 10 minutes gives us 492 residues with 479 docked into the sequence (we expect 543 residues, the PDB entry contains 498 residues):