Help

Each assembly is represented by a thumbnail image in fat ribbon representation with different subunits in different colors (1st column). Also a 2-dimensional graph diagram of the assembly is shown, with nodes being the chains and edges the interfaces between them (2nd column). Distinct molecular entities and distinct interfaces are depicted with different colors. Next appear the macromolecular size, stoichiometry (the successive letters represent different molecular entities, not chain ids) and point group symmetry (Cn for cyclic symmetries, Dn for dihedral symmetries, T for tetrahedral, O for octahedral and I for icosahedral).

Note that disjoint assemblies will be shown separated by commas in the size, symmetry, and stoichiometry columns. Disjoint assemblies are those where not all components of the crystal form a single assembly but instead several disjoint ones, e.g. a crystal containing 2 protein entities will always contain a disjoint assembly formed by 2 independent monomers (one for each entity).

The prediction column provides the predicted assembly (marked as BIO), while all others are marked as XTAL. The predicted assembly corresponds to that with the highest calculated probability, based on the scores of the interfaces that form it. The probability values for each assembly appear next to the call. An estimation of the prediction confidence is provided with star icons: golden star for high confidence, gray star for medium confidence, no star for low confidence.

The last column shows how many interfaces compose the assembly. Clicking on the button, the interfaces table is shown with only those interfaces belonging to the assembly. The view can be reset to show all interfaces by clicking on the close icon in the tab title.

Each of these indicators have predefined score thresholds to produce one of the calls:

• BIO, the interface is biologically relevant
• XTAL, the interface is only a crystal lattice contact
• NOPRED, there is not enough information available to make a decision (usually not enough sequence data)

You can see the scores for each of the indicators next to the bio/xtal/nopred labels.

The two scores are used to calculate a final score and a probability of the interface being biologically relevant (1 being certainly biological, 0 certainly crystal contact). The call and the probability appear in the "Final" column. BIO will mean that the probability is above 0.5 and XTAL that the probability is below 0.5. This is the final prediction column and what you need to look at first. An estimated confidence level for the prediction is depicted with stars, golden star for high confidence, gray star for medium confidence, no star for low confidence.

The other columns in the interface table correspond to a few important parameters describing the interfaces: the two chain codes of the partners (e.g. "A+B"), the Buried Surface Area upon interface formation (interface sorting is based on this value), the icon of the crystal operator used to generate the second partner of the interface

The operators are represented as icons to show at a glance what kind of crystallographic symmetry is present at the interface. The actual full algebraic operator (e.g. "-X+1,Y-1/2,-Z") can still be seen by hovering the mouse over the icon. The icons used for the operators are mostly the standard ones found in crystallographic tables: the identity operator (i.e. an interface in the asymmetric unit), a crystal translation (integer) without rotation, a re-centering translation without rotation, a 2-fold axis, a 2-fold screw axis, a 3-fold axis, a 3-fold screw axis, a 4-fold axis, a 4-fold screw axis, a 6-fold axis, a 6-fold screw axis.

For the rare cases where a protein is crystallized in non-chiral space groups (e.g. racemic mixtures) there are additional operators: an inversion centre, a mirror plane, a glide plane, an improper 3-fold axis, an improper 4-fold axis, an improper 6-fold axis.

Some of these operators lead to the formation of infinite interfaces if occurring between two crystallographically-related copies of the same molecule (e.g. A+A). This happens for both the pure translations and for any of the screw rotations and it is generally a very strong indication of a crystal contact. In those cases, we color the operator icon in red. The final call does not take that information into account, but this is very important for the enumeration of valid assemblies in the crystal shown in the assemblies table.

The thumbnails in the Assembly and Interface tables give a visual cartoon representation of the assemblies and interfaces. By clicking on them one gets an interactive 3D view with the Mol* viewer. In the interface view the two protomers are represented as cartoons with interface residues also shown as sticks. Core residues from both protomers are shown in two different shades of red. The sequence entropy values for each residue are written as b-factors in the .cif file.