***  For details on how to be removed from this list visit the  ***
***          CCP4 home page http://www.ccp4.ac.uk         ***


Many thanks to all of those who emailed me (some off-list) about my
recent request: the best-practice methods to enhance "missing domain"
density in molecular replacement solutions by solvent-flattening.  The
specific case I had in mind was at a modest resolution of ~3 Angstrom
although I did not specify that in my original post.


Bart Hazes pointed out that not much might be expected to be gained from
pure solvent-flattening alone, since the phases calculated from the
model feature largely flat bulk solvent in the first place.  And I
cannot disagree with him there, although I was hoping to get a little
boost from histogram matching etc.

Dirk Kostrewa and Petro Roversi suggested using BUSTER-TNT and provided
this reference: Blanc, Roversi, Vonrhein, Flensburg, Lea and Bricogne,
Acta Cryst. (2004). D60, 2210-2221 "Refinement of severely incomplete
structures with maximum likelihood in BUSTER-TNT."

Megan Maher, Anastassis Perrakis suggested using RESOLVE with the Prime
and Switch method.  (This seems especially straightforward to use with
an example at this URL:
http://www.solve.lanl.gov/html_resolve/resolve_examples.htm and example
script here:
http://www.solve.lanl.gov/html_resolve/resolve_sample_scripts.htm#ps)

Sheena McGowan and Peter Meyer indicated that PIRATE was giving them the
best results.

Bill Scott suggested using EDEN with an explicitly generated solvent mask

Phoebe Rice emphasized manual mask definitions, sigmaA maps and phase
extension from lower resolution.

Anastassis Perrakis also suggested ARP/wARP at better than 2.0 Angstrom,
as well he might since I was having success with that program only this
week in an unrelated SeMet MAD case (and not for the first time).

And now that most of the synchrotron data is out of the way, I hope to
try a few of these methods.

Thanks again to all those who emailed me,
Phil Jeffrey
Princeton, NJ


Reply via email to