On Feb 20, 2009, at 2:58 PM, James Holton wrote:
I made a movie demonstrating the effect of leaving out low-res data
on an electron density map once:
http://bl831.als.lbl.gov/~jamesh/movies/index.html#lores
Very cool movie.
Many thanks to those who answered my original question by private email.
After reflecting for a while, I realize I should have asked a
slightly different question.
James, your movie shows nicely that one can still have pretty good
density even with data below 6A omitted. I presume there is still a
reasonable-looking envelope at that point, though I can't tell from
looking. The resolution limit for that set is 1.5A.
Now here's my question: suppose you increase the resolution to 1.0A.
Can you then cut even more low resolution (say down to 5A) before the
density goes bad?
I am used to thinking that the envelope shape comes entirely from the
low-resolution data. Indeed, a SAXS pattern with
data ranging from say 200A to 15A will generate a nice envelope. It
occurs to me, however, that this is not the whole story.
If you have the molecular coordinates, you have the envelope; and
there is lots of high-precision molecular coordinate information
encoded in the higher resolution shells. Conventional processing
techniques may not work in that regime, but it seems that the
information is there.
Perhaps someone with expertise in ab initio methods can comment, but
I recall hearing that the molecular structure is actually over-
determined if one had access to the full molecular transform, so we
are always only using bits and pieces of it in crystallography. But
to what extent that can, in principle, be abused by leaving out low
resolution I would like to know.
Richard Gillilan
MacCHESS
