Hi,
1. do you know for sure that you have useful
Se incorporation levels (i.e. by MS)?
2. there are minimal distance constraints
on how close heavy atoms may be (one could use general chemistry knowledge to figure
them out). Two actual Se atoms should never be as close as 1A apart. The sites
can appear to be this close if you’re dealing with disorder (but then
occupancies will be a giveaway) or with a twinned crystal (in both cases the actual
atoms aren’t really 1A apart but they appear to be) – and in the
latter case all bets are off. Did you thoroughly check for twinning? Based on
your knowledge of the enzyme in question – is there anything funky like a
possible sulfur-iron cluster, metal ion coordination, etc.?
3. I am not sure how to interpret your
comments regarding resolution and anomalous signal quality – 1.6A is
certainly nice (assuming that you and I use the same terms to describe ‘good
quality’) and to be honest useful anomalous signal to 2.2A is also quite
nice.
4. Returning to the choice of space group –
are you sure it’s P21? There is such a thing as a P2 protein crystal
although the instance of that SG is quite rare (95 entries in the current PDB)
in the protein world. Again, I would take another close look at twinning
possibilities.
Good luck,
Artem
“Nothing is built on stone; all is
built on sand, but we must build as if the sand were stone”
Jorge Luis Borges
From: CCP4 bulletin
board [mailto:[log in to unmask]] On
Behalf Of yang li
Sent: Tuesday, July 28, 2009 10:10
PM
To: [log in to unmask]
Subject: [ccp4bb] Deal with close
heavy atom sites?
Dear all,
I have a 1.6A data in
P21 spacegroup whith good quality, wich is assumed to contain 4
Set-Met. But the patterson map
appears strange. There is a very big heavy peak in the harker
section, and seems like an irregular ellipse, while there are small
peaks very near(almost connect) to this big peak and the origin,
say about 0.05 in fraction, looks like cross peaks.It looks like the
heavy atoms are very close if there are more than one. I tried several
programs and many heavy atom coordinates sits, but none
of them can solve it. Shelxd got a set of sites looks best with set the
minimum distance to 1A, since the patterson map from it
looks like the experimental patterson map, though not very much. But
with this HA sites can get nothing meaningful. By the way,
it is a small molecule with no more than 150 residues, one
molecule/ASU. According to shelxc the signal is not strong, can only
be used to aroud 2.2A.
Any suggestions would be
appreciated!