Hi Fellows,
an interesting story about the combined effects of phase bias in MR and bad
luck (or lack of skepticism)
in protein preparation just appeared:
http://hmg.oxfordjournals.org/content/early/2016/08/29/hmg.ddw298
These cases may not be quite so rare: A fellow of mine recently collected an
exciting data set
of his new fusion-tagged protein, which promptly solved with a MBP dimer....
The risk of low resolution data and poor MR models generally seems to be
underappreciated.
Particularly around 3A and slightly lower, the temptation to press on
because the biased maps
don't look bad enough to give up right away (but without much indications of
meaningful
corrections) can prompt wishful manipulation of other refinement parameters
until some apparently acceptable outcome results...
Maybe it would be a useful addition to our powerful brute force MR programs
to have a little
button (or default) that says 'Run against data base of known co-purifying
contaminants'
Anyone interested?
Cheers, BR
------------------------------------------------------
Bernhard Rupp
http://www.hofkristallamt.org/
[log in to unmask]
+1 925 209 7429
+43 767 571 0536
------------------------------------------------------
Many plausible ideas vanish
at the presence of thought
------------------------------------------------------
|