Occupancies

EU name: 1QU9

(Date: Aug 24 2016 1QU9 )

1QU9

We all know that if an atom sits at a special position, i.e. exactly on a symmetry axis, then you should reduce its occupancy in the PDB file by the exact same factor as the multiplicity of that symmetry axis. 

JRNL        AUTH   K.VOLZ
JRNL        TITL   A TEST CASE FOR STRUCTURE-BASED FUNCTIONAL
JRNL        TITL 2 ASSIGNMENT: THE 1.2 A CRYSTAL STRUCTURE OF THE
JRNL        TITL 3 YJGF GENE PRODUCT FROM ESCHERICHIA COLI
JRNL        REF    PROTEIN SCI.                  V.   8  2428 1999

In this file we find a series of water molecules with occupancies that suggest that they are located at special positions. In the box below I have taken only the waters from 1QU9 that have an occupancy other than 1.0. 

HETATM 2986  O   HOH A 345      36.123  72.223  49.938  0.50 35.86           O
HETATM 3154  O   HOH B 545      72.223  49.938  36.123  0.50 35.86           O
HETATM 3313  O   HOH C 745      49.938  36.123  72.223  0.50 35.86           O
HETATM 2902  O   HOH A 256      36.121  69.677  72.242  0.50 18.29           O
HETATM 3068  O   HOH B 456      69.677  72.242  36.121  0.50 18.29           O
HETATM 3229  O   HOH C 656      72.242  36.121  69.677  0.50 18.29           O
HETATM 2928  O   HOH A 284      39.241  72.228  72.228  0.50 23.81           O
HETATM 3095  O   HOH B 484      72.228  72.228  39.241  0.50 23.81           O
HETATM 3255  O   HOH C 684      72.228  39.241  72.228  0.50 23.81           O
HETATM 2950  O   HOH A 306      47.693  80.499  51.736  0.50 27.31           O
HETATM 3117  O   HOH B 506      80.499  51.736  47.693  0.50 27.31           O
HETATM 3277  O   HOH C 706      51.736  47.693  80.499  0.50 27.31           O
 
HETATM 3014  O   HOH B 330      43.028  43.028  43.028  0.33 32.19           O
HETATM 3139  O   HOH B 530      43.028  43.028  43.028  0.33 32.19           O
HETATM 3169  O   HOH B 730      43.028  43.028  43.028  0.33 32.19           O
HETATM 3007  O   HOH B 206      45.290  45.290  45.290  0.33  8.18           O
HETATM 2998  O   HOH A 406      45.290  45.290  45.290  0.33  8.18           O
HETATM 3166  O   HOH B 606      45.290  45.290  45.290  0.33  8.18           O
HETATM 3010  O   HOH B 276      54.593  54.593  54.593  0.33 21.40           O
HETATM 3088  O   HOH B 476      54.593  54.593  54.593  0.33 21.40           O
HETATM 3168  O   HOH B 676      54.593  54.593  54.593  0.33 21.40           O

The CRYST and SCALE cards indeed indicate the existence of 2-fold and 3-fold axes. 

CRYST1   72.220   72.220   72.220  90.00  90.00  90.00 P 2 2 2      12
SCALE1      0.013847  0.000000  0.000000        0.00000
SCALE2      0.000000  0.013847  0.000000        0.00000
SCALE3      0.000000  0.000000  0.013847        0.00000

So I can somewhat understand some of the the waters with occupancy 0.5. Some other waters with occupancy 0.5, however, seem at a first glance to not be exactly on a 2-fold axis. The atoms with occupancy 0.33 indeed seem positioned exactly at a 3-fold axis as the view below down that 3-fold axis seems to suggest. If you put a water with occupancy 0.33 on a 3-fold axis, there should not be three instances of that water, not even with different chain identifiers.

I am not such a hero with spacegroups and spacegroup symmetry, but I am a bit suprised this spacegroup is called P222 and not P23, but that is a different story.

A second note is that the spacegroup once was P23, but that the PDB expanded the file and made if P222, so it doesn't seem overly unlikely that the PDB is to blame for this problem.

The resolution of 1QU9, by the way, is 1.2 Ångström, so there should be enough data to do things well.

Waters with non-1.0 occupancy shown in a wireframe cell of 1QU9.

EU name: 1VR1

(Date: Aug 24 2016 1VR1 )

1VR1

Occupancies other than 1.0 are used when atoms are found at different locations either at different times or in different asymmetric units. Often though, they are also used to indicate that atoms are missing, invisible, or otherwise doubtful. 

JRNL        AUTH   R.J.DEKKER,A.EICHINGER,A.A.STOOP,W.BODE,
JRNL        AUTH 2 H.PANNEKOEK,A.J.G.HORREVOETS
JRNL        TITL   THE VARIABLE REGION-1 FROM TISSUE-TYPE PLASMINOGEN
JRNL        TITL 2 ACTIVATOR CONFERS SPECIFICITY FOR PLASMINOGEN
JRNL        TITL 3 ACTIVATOR INHIBITOR-1 TO THROMBIN BY FACILITATING
JRNL        TITL 4 CATALYSIS: RELEASE OF A KINETIC BLOCK BY A
JRNL        TITL 5 HETEROLOGOUS PROTEIN SURFACE LOOP
JRNL        REF    J.MOL.BIOL.                   V. 293   613 1999

In the file 1VR1, though, we find some funny zero occupancies: 

ATOM    996  N   GLU H  97A     28.200 -13.994  26.148  1.00 42.09           N
ATOM    997  CA  GLU H  97A     27.495 -14.593  27.279  1.00 42.24           C
ATOM    998  C   GLU H  97A     26.054 -14.119  27.395  1.00 47.57           C
ATOM    999  O   GLU H  97A     25.064 -14.814  27.268  1.00 49.99           O
ATOM   1000  CB  GLU H  97A     28.231 -14.164  28.561  1.00 44.76           C
ATOM   1001  CG  GLU H  97A     28.577 -13.500  28.894  0.00 37.87           C
ATOM   1002  CD  GLU H  97A     29.575 -12.878  30.251  1.00 58.21           C
ATOM   1003  OE1 GLU H  97A     28.669 -13.209  31.495  0.00 38.48           O
ATOM   1004  OE2 GLU H  97A     30.490 -13.001  30.300  0.00 38.47           O
 
HETATM 2396  O   HOH H 313      28.850 -13.227  29.373  1.00 81.71           O

I added one water to this box. One doesn't need a calculator to see that this water resides 'inside' the glutamate. The distance of this water to the Cγ and Cδ, respectively is 0.6 Ångström  and 1.2 Ångström.

1VR1 as stick model. Red spheres are waters. The two yellow spheres are waters that reside inside residues (probably there are more, the 'other' yellow one I spotted by eye accidentally and coloured to show that the one water inside Glu-97A is not an individual incident). Glu-97A is shown in magenta (near the bottom).