Dear all
 
Thank you for all the helpful suggestions:
 

Artem Evdokimov:
 
What’s the resolution and VM of your crystal? It’s very unlikely that you have only 26 water molecules around �C but you may not be able to locate most of the solvent present, if the resolution is too low and/or the disorder too high…
 
With the parameters you indicated, you have two likely possibilities �C either a monomer with 70% solvent content, or a dimer with 40% solvent content. A trimer would mean 13% solvent content �C way too low to be likely.
 
 
 
 
 
Edwin Pozharski:
 
Check this paper:
 
*Author(s):* Carugo, O; Bordo, D
*Title:* How many water molecules can be detected by protein
crystallography?
*Source:* ACTA CRYSTALLOGRAPHICA SECTION D-BIOLOGICAL CRYSTALLOGRAPHY,
55: 479-483 Part 2 FEB 1999
 
It does, of course, describe how many water molecules people add to
their structures, not how many can really be located. In my estimate
(based on the above paper) you should have around 150 water molecules
per monomer, but again, this is just average expectation. maybe your
B-factors are higher than usual for a structure of 2.5A resolution and
then you see fewer waters.
 
Your solvent content is about 40% - somewhat lower than normal.
Accordingly, you may have more of the protein surface buried in contacts
and therefore fewer waters. But you can't say that waters aren't there -
they are simply disordered and you can't see them.
 
I do not know of any paper that would address this question. I would say
this B-factor is somewhat higher than usual. I can tell you that out of
835 structures from the PDB with reported overall B-factor and
resolution of 2.5A only 191 (about 23%) have reported overall B-factor
of more than 47.
 
I would not worry much about this issue. The number of waters people
report depend on number of things - maybe you are just being too
conservative with your water picking procedure.
 
 

Petrus H Zwart:
 
FYI, the cell you gave is pseudo C2221 (see below). I however doubt it
has anything to do with the water issue.
%iotbx.ehms --unit_cell=40,75,99,90,101,90 --space_group=P1211 --
max_delta=2.5
 
A summary of the constructed point group graph object is given below
====================================================================
 
----------------------
Input crystal symmetry
----------------------
Unit cell:  (40.0, 75.0, 99.0, 90.0, 101.0, 90.0)
Unit cell volume:  291543.273484
Space group:  P 1 21 1
 

--------------------------
Lattice symmetry deduction
--------------------------
Niggli cell:  (40.0, 75.0, 99.0, 90.0, 101.0, 90.0)
Niggli cell volume:  291543.273484
Niggli transformed input symmetry:  P 1 21 1
Symmetry of Niggli cell:  Hall:  C 2 2 (x+y,z,2*x)
 

All pointgroups that are both a subgroup of the lattice symmetry and
a supergroup of the Niggli transformed input symmetry wil now be listed,
as well as their minimal supergroups/maximal subgroups and symmetry
operators that generate them.
For each pointgroup, a list of compatible spacegroups will be listed.
Care is takebn that there are no sysmetatic absence violation with the
provided input spacegroup.
 
------------------------
Vertices and their edges
------------------------
 
Point group   Hall:  C 2 2 (x+y,z,2*x)   is a maximal subgroup of :
  * None
 
Point group   P 1 2 1   is a maximal subgroup of :
  * Hall:  C 2 2 (x+y,z,2*x)
 
 
 
-------------------------
Transforming point groups
-------------------------
 
From P 1 2 1   to  Hall:  C 2 2 (x+y,z,2*x)  using :
  *  -h,-k,h+l
 
 
 
----------------------
Compatible spacegroups
----------------------
 
Spacegroups compatible with a specified point group
**and** with the systematic absenses specified by the
input space group, are listed below.
 
Spacegroup candidates in point group Hall:  C 2 2 (x+y,z,2*x):
  * C 2 2 21  40.00 194.37 75.00 90.00 90.00 90.00
 
Spacegroup candidates in point group P 1 2 1:
  * P 1 21 1  40.00 75.00 99.00 90.00 101.00 90.00
 
 
 
 
 
Ramanathan NATESH:
 
     I would assume your resolution is on lower side, less than 3 Ang ?
If so .. you will obviously find less waters whatever may be the duration
of crystallization.
 
 
 
Vaheh Oganesyan:
 
The number of water molecules depends on resolution of your data and hydrophobicity of the protein, and not on how long it took to crystallize.
 
 
 
 
 

LiangZhang
2006-06-08

Reply via email to