3did: a catalog of domain-based interactions of known three-dimensional structure

Abstract

The database of 3D interacting domains (3did, available online for browsing and bulk download at http://3did.irbbarcelona.org) is a catalog of protein-protein interactions for which a high-resolution 3D structure is known. 3did collects and classifies all structural templates of domain-domain interactions in the Protein Data Bank, providing molecular details for such interactions. The current version also includes a pipeline for the discovery and annotation of novel domain-motif interactions. For every interaction, 3did identifies and groups different binding modes by clustering similar interfaces into 'interaction topologies'. By maintaining a constantly updated collection of domain-based structural interaction templates, 3did is a reference source of information for the structural characterization of protein interaction networks. 3did is updated every 6 months.

Figure 1.

Growth of 3did throughout its four releases. The colored bars represent the number of DDI pairs with only intrachain structural templates (dark green), only interchain templates (medium green) and both types of templates (light green). Bar plots refer to the y-axis on the left. The lines represent the growth in the number of structures (dark orange) and chains (light orange) in the PDB. The yellow line represents the number of domain–domain structural templates in 3did (i.e. the number of redundant structural instances of DDI). Line plots refer to the y-axis on the right.

Figure 2.

Overview of the DMI discovery pipeline. The main steps of the DMI discovery pipeline are outlined, with filtering steps to remove spurious hits. Details can be found in (33).

Figure 3.

Views available in 3did. 3did provides four views to browse the data contained in the database: the Domain view (A), the Motif view (B), the PDB view (C) and the Interaction view (D).

Figure 4.

Browsing 3did. (A) Interactive view of the DDI and DMI network involving a particular domain. In orange are the domains while the motifs are in green. By clicking on any node or edge you are redirected to the page showing details about the corresponding domain, motif or interaction. (B) Interactive view of the domains and motifs in a PDB file. It shows the domain composition of the different chains (clustered on identical domain composition) as well as the motifs present in the chains. Lines connect domains and motifs that are interacting. Both nodes and lines can be clicked in order to visualize the details of the corresponding domain, motif or interaction. The CytoscapeWeb-based network visualizations (A and B) require a Flash plug-in to be installed in the browser to run. (C) Browse by GO term. A new tree view in the ‘Browse’ page allows searching for all the domains that are annotated with a specific GO-term.