Hello Doug,
I understand your arguing, although I do not understand why Fo contains a
contribution from the model(led structure). My idea was only supposed to
be a backup in case the apo-form had not been measuredcor could not be
used. I agree the Fo(complex)-Fo(apo) gives a much more realistic result
than the MR-solution I suggested.
Tim
--
Tim Gruene
Institut fuer anorganische Chemie
Tammannstr. 4
D-37077 Goettingen
GPG Key ID = A46BEE1A
On Thu, 18 Jun 2009, Doug Ohlendorf wrote:
Tim,
Not really. Fo's can be viewed as having 3 parts. 1) F from our modeled
structure, 2) F from what we can't model. This can be bulk solvent, partial
or multiple occupancies in low resolution structures, thermal anisotropy ,
etc. 3) F from random errors in measuring data. The first F is what we deal
with. The third F we can minimize by collecting high quality data. The
second F is usually crystal/space group dependent. It arises from
crystallization conditions, crystal morphology and intermolecular contacts.
Data from isomorphic crystals should have similar second F's so map (from
first F's) is cleaner.
Doug
-----Original Message-----
From: CCP4 bulletin board [mailto:[email protected]] On Behalf Of Tim
Gruene
Sent: Wednesday, June 17, 2009 4:39 PM
To: [email protected]
Subject: Re: [ccp4bb] Difference Map images
I know that sometimes Fo(complex)-Fo(apo) cannot be done because of
nonisomorphism. We've had a lot of success with this with the dioxygenases
because there is no large scale alteration in the active site. As for the
technique itself, Brian Matthews drilled this into me when I was a postdoc
in his lab.
Wouldn't - in the case of non-isomorphsim - a molecular replacement with
the apo-form come closest to a Fo(complex)-Fo(apo) map?
Just a thought.
Tim
