As much as I hate to see the CCP4BB used for shameless self-promotion of
articles, I did just publish a "Beginner's Guide to Radiation Damage":
http://journals.iucr.org/s/issues/2009/02/00/issconts.html
My answers to Paul and Michael's posts would be cut-and-pasted from it,
so I think it appropriate to refer to it here.
I even made the cover! And I hope that (after looking at the cover) all
you BBers will forgive me just this once as this was clearly a
once-in-a-lifetime opportunity. Or, at least, I HOPE only once.
-James Holton
MAD Scientist
Paul Swepston wrote:
> Michael:
>
> I attended the CCP4 study weekend in January and there were many discussions
> about radiation damage. In your case I wonder if recollecting the data with
> a lower over all dosage might get rid of the problem. Either collect with
> an attenuated beam or collect the data twice as fast. Someone with more
> experience should chime in on this suggestion.
>
> Paul
>
> -----Original Message-----
> From: CCP4 bulletin board [mailto:[log in to unmask]] On Behalf Of
> Michael Weyand
> Sent: Friday, February 27, 2009 6:25 AM
> To: [log in to unmask]
> Subject: [ccp4bb] Refmac5 - high res (1.2A) refinement problem
>
> Dear experts,
>
> here is a Refmac question (or at least a refinement question):
>
> the case: data set was collected at the synchrotron with high intensity or
> better high total dose. Data set scales to 1.2A resolution.
> Refinement with refmac5_0066, using hydrogens, anisotropic B-factors,
> automatic weight (actually 17.9), 235 amino acids, approx. 450 waters,
> several ions and buffer/crystallization compounds, a hell of second and
> third side chain conformations, present R-fac 12.6, free-R 15.6.
>
> I think I'm seeing a (partial) backbone break (negative DelFwt density for
> backbone atoms at 3.5sig within coot) due to radiation damage in a surface
> loop. I see clear missing COOH- and Guanidinium-groups in other parts of the
> model!!
>
> As a consequence, the protein adopts a second (backbone) conformation (clear
> positive DelFwt at 4.0sig) for an alpha-helix. But some of the second
> conformation residues are not visible.
> Summary:
> conformation A (aa206-219) ~ 70% occupancy, conformation B (aa212-219) ~ 30%
> occupancy, no density (flexibility??? / deletion ???) for aa206-211.
>
> Refmac always link aa212, conformation B to aa211, conformation A. But, the
> CA distance between Calpha atoms of aa212 is / should be about 5A.
>
> How to address the "non-restraint" between these amino acids, or vice versa
> to address a (partial) chain break?
>
> Furthermore I see "partial" waters at a (second) side chain
> conformation/position. Is it also possible to release the restraints for
> these waters to atoms of the second side conformation?
>
> It is really important to do it right, since the second conformation
> residues change the active site with a bound ligand.
>
> Every hint or work-around is highly appreciated.
> May be I have to use SHELXL?
>
> Regards
> Michael
>
>
>
>
|