The Ensemble geometry plot shows how the protein's geometrical properties vary along its sequence. This gives a visualization of which regions appear to have consistently poor geometry (perhaps because they are poorly defined) and which have more normal geometry.

The properties plotted are:-

**Graphs a-b: Optional properties**The first two graphs at the top of the page, can be selected from

**3**possibles by the user. The two default graphs, which are plotted when you first run**PROCHECK-NMR**, are the first two of:-- 1. Absolute deviation of the residue's
**chi1**torsion angle from the nearest**"ideal"**value. - 2. Absolute deviation of
**omega**torsion angle from the**"ideal"**. - 3. Absolute deviation of
**zeta "virtual"**torsion angle (defined by the atoms**C**) from the*alpha*-N-C-C*beta***"ideal"**.

For each graph, the values for each residue in each model are plotted as individual crosses. The mean and standard deviation values for a given residue are indicated by the circle and bars, respectively. The model whose value is the highest is indicated by the model-number printed above the highest mark.

The dashed line corresponds to

**2.0**standard deviations from the**ideal**, so an excess of points above the line suggest possible problems with the geometry.- 1. Absolute deviation of the residue's
**Graph c: RMS deviation from mean coordinates**The

**RMS deviations**plot shows a histogram of the*rms*deviations from the mean coordinates of the**main-chain**atoms (**black**bars) and the**sidechain**atoms (**grey**bars). The mean coordinates are currently calculated simply by averaging each atom's coordinates across the whole ensemble.**Graph d: Secondary structure & average estimated accessibility**The

**secondary structure plot**shows a**schematic**representation of the**Kabsch & Sander (1983)**secondary structure assignments. The key just below the picture shows which structure is which. Beta strands are taken to include all residues with a Kabsch & Sander assignment of**E**, helices corresponds to both**H**and**G**assignments, while everything else is taken to be random coil.The shading behind the schematic picture gives an

**approximation**to the**residue accessibilities**. The approximation is a fairly crude one, being based on each residue's**Ooi**number (Nishikawa & Ooi, 1986). An**Ooi**number is a count of the number of other**C**atoms within a radius of, in this case,*alpha***14Å**of the given residue's own**C**. Although crude, this does give a good impression of which parts of the structure are buried and which are exposed on the surface. Future versions of*alpha***PROCHECK**will include an accurate calculation of residue accessibility.**Graph e: Sequence & average estimated accessibility**The next section shows the

**sequence**of the structure (using the 20 standard amino-acid codes) and a set of markers that give a schematic representation of the**estimated accessibility**. The latter is calculated as above, but here is show such that the darker the symbol the higher the accessibility.**Graph f: Circular variances**The dials give a schematic representation of each residue's

**circular variance**values for its**phi**,**psi**,**chi-1**and**chi-2**angles, and for its**phi-psi**and**chi1-chi2**combinations. The larger the black segment on the dial, the higher the**circular variance**, and hence the wider the**spread**of the corresponding dihedral angle distribution.Regions with many black, or near-black, dials correspond to regions where there is a large variability in the residue's tordion angles across the models in the ensemble. These may correspond to highly mobile or poorly defined regions such as loops, or may need further investigation.

**Graph g: G-factors**The shaded squares give a schematic representation of each residue's

values for its*G*-factor**phi-psi**and**chi1-chi2**and**chi-1**dihedral angles. (Note that the**chi-1***G*-factors are shown only for those residues that do not have a**chi-2**, and hence no**chi1-chi2***G*-factor).Regions with many dark squares correspond to regions where the dihedral angles are "

**unusual**", as defined by a low (or negative)*G*-factor. These may correspond to highly mobile or poorly defined regions such as loops, or may need further investigation.

The main options for the Ensemble geometry are:-

- Different graphs can be plotted in the
**2**main graph areas at the top of the plot. - The plot can be in colour or black-and-white.

These options can be altered by editing the parameter file,
**procheck_nmr.prm**, as described here.