0%

​

Share this page with:

Structures and AlphaFold subpages

In this part of the tutorial we will look at the structure information you can find in Pfam. Continuing with our example PF20141, the structure information is accessible from the structures icon at the top or from the list on the left.

Structures subpage

This subpage provides the list of experimentally determined structures found for protein sequences included in this entry, from the PDB database (Figure 11). Each of them has a link to the corresponding structure page in Pfam and the option to view the structure in the same page (‘View 3D structure’) or in InterPro (the link will redirect you to the InterPro page). Additionally, a link to the protein sequences (i.e. BRI1_ARATH) redirects you to the protein page in Pfam.

Figure 11. Structures subpage of PF20141, listing all the PDB structures for protein sequences included in this entry.
In the Pfam entry page for PF20141 accessed before, go to the ‘Structures‘ subpage to see the PDB structures available for this entry. You can also click on the ‘BRI1_ARATH’ accession in the table to access the corresponding Protein page.

Let’s have a look at the structures for the island domain PF20141, for example 3rgx. In the interactive structure viewers (Figure 12), we can see the characteristic horseshoe structure of LRRs with the island domain inserted within LRRs and the type of arrangement it has (alpha-beta structure). As we already know from the description and sequence mapping, we can see and confirm that the island domain is inserted between LRRs (red circle in the left image or in blue in the right one in Figure 12). By hovering over different parts of the representation, it is possible to see which amino acids are in the structure and their position.

Figure 12. Two views of 3rgx structure. On the left, the structure chain all in the same colour, and on the right the island domain is coloured in blue.

If we click on the structure PDB ID, we will find general information about the structure and links to other databases, such as PDBe (Figure 13). Additionally, from the left-hand side menu you can access the domain organisation in which the structure is found, the sequence mapping (positions of each domain found in the sequence with this structure) and a link to the structure viewer in Pfam.

Figure 13. PDB structure page of 3rgx.

AlphaFold structures subpage

PF20141 also has an AlphaFold structure predictions. These predictions are generated through a collaboration between DeepMind and EMBL-EBI using Artificial Intelligence based on full length protein sequences. In Figure 14 we can see the list of sequences for which a structure prediction has been generated.

Figure 14. AlphaFold structures prediction subpage for PF20141.

When clicking on “View 3D Structure” for any of the sequences in the list, you are redirected to the AlphaFold Structure subpage of the corresponding protein page.

For example, if you click on the link for viewing A0A0R0KC24, the predicted structure is displayed as shown in Figure 15. The island domain is highlighted with a red circle. As the AlphaFold predictions use the full length of the sequence, they allow us to see how the relative positions of the domains are, their possible interactions and how the protein arranges in space.

If you remember the domain organisation of the PF20141 entry and the BRI1_ARATH protein, apart from the LRRs and the island domain, they also have a kinase domain, which is visualised in the AlphaFold prediction (the alpha-helical/beta-strand domain at the bottom of the second image or at the top of the third one in Figure 15).

Figure 15. AlphaFold structure prediction for A0A0R0KC24, view from different angles.

In the next section of this tutorial you will find another type of structure prediction: the trRosetta models.