PDBnet Bird's eye view of network in structurome

Tutorials

PDB chains, Domains, Clusters, Interactions and Networks

   The PDB complexes contain the units of structures called "chains" corresponding to molecules such as proteins and nucleic acids (the example shown here is PDB ID=2ata). The chains consist of "domains" as the units of structure and function. The chains and domains forming "clusters" with homologous sequences interact with others within the complex structures. Those interactions then form the "networks" of clusters. PDBnet is intended to integrate the information of these entities.

The example shown here is 2ata

To RCSB PDB : 2ata


Searching PDBnet

   Search will begin by entering PDB ID, Chain ID, UniProt ID or Gene name in the search box.


   When PDB ID: 1ydp is entered, the following page with the list of PDB chains will be displayed:


   Search will begin by entering PDB ID, Chain ID, UniProt ID or Gene name in the search box.


Advanced Search page

   You can search PDBnet by "PDB Code", "Molecule", "Genus", and "Length", "Chain Type", and "Keyword". For instance, when '1gmj' is searched as PDB code, the following result is obtained.

Result : The example shown here is PDB Code=2ata

Clicking PDB-chain 1gmjA leads to Chain information page.


Chain information page

   This page shows the major information about the PDB chain such as molecule name, synonym, source and length. The next panel shows the structure of Asymmetric Unit and Biological Assembly by Jmol. Shown on the left is the interaction network among chains within the complex structure. Clicking each edge will pop up a new window showing the Domain and Non-domain Interaction Network between the chains (P00533 (EGFR) in this example).


Cluster information page

    This page displays query protein cluster (shown in yellow) and the network of interacting clusters. White nodes represent protein clusters and orange nodes correspond to nucleic acids. Red edges represent 'hard links' (physical contacts within 5 angstrom). The protein clusters are represented by a UniProt ID. Detailed information can be displayed when cursor is placed on each node. Clicking each edge will display a link below the graph, and clicking the link leads to the "Domain information" page (see the later section) in another window. The next 1-D panel shows all the PDB chains (shown in blue) and Pfam domains (violet) involved in this query cluster (clicking on the blue or violet region on the left will display the list on the right). The next graph shows a plot of amino-acid composition, amino-acid conservation score, or B-factor of the query protein (P00533 in this example). The last tables show the detailed information about the PDB chains in the query cluster and the interacting clusters.

The drawing of network is done by Cytoscape Web.
Cytoscape Web : http://cytoscapeweb.cytoscape.org/


Biochemical PPI information page

   This page shows protein-protein interaction (PPI) information obtained by biochemical experiments. Dashed edges represents interactions only from biochemical experiments, and the thickness of edge lines are set according to the reliability of interaction. The reliability score was evaluated according to the type of experiment: high score for low throughput, and low score for high-throughput experiments. Also, the score is raised if the same interaction was verified by different experiments. The red edges indicate that the interactions are observed in both the structural and biochemical PPI. The network is shown for every 30 interactions to prevent the graph from becoming overcrowded. The interacting clusters can be switched to those with 50% sequence similarity by choosing from the list menu below the graph (structural interactions will be shown in orange edges). When "DDI prediction network" is selected in the list menu, those interactions predicted according to the domain-domain contacts verified by experimental structures will be highlighted by violet lines.

The example shown here is ClusterNo=P00533


Domain information page

   Clicking each edge of the interaction network in the "Cluster information" page will display a link below the graph, and clicking the link leads to this page showing interactions between protein domains (Pfam domains) as well as chain-chain interactions. This figure shows an example of EGFR (P00533) in yellow. Users can select a particular chain-chain pair at the top menu. The interacting chains are mapped on each sequence (top panel). Domains in each chain are represented by square nodes and non-domains by triangles. You can see which domains and non-domains are interacting with each other. The corresponding 3-D structure plotted by Jmol is shown on the right. Clicking the domain nodes on the left will highlight the corresponding structure in orange, and clicking the edges will show the interacting regions in the structure in violet. The panel below the Jmol structure displays the experimentally-verified conserved motifs (taken from PROSITE) in domains and non-domains. Pressing the "Click" button at the last Jmol column will highlight the corresponding motif mapped onto the Jmol structure in red. The detailed information of the domains and non-domains is shown as 1-D panel at the bottom. Red regions represent domains with Pfam ID and grey regions are non-domains. Those amino acid residues involved in the interactions are shown in red bars.


   For domain interaction information, see the "Detailed interaction information page" section.


Network information page

    The network information page can be accessed by clicking "Network" label at the top menu in "Chain information" or "Cluster information" page. The network including the query protein is shown as a diagram in the same way as in the Cluster diagram. Nucleic acids are designated in orange color. Each node is hyperlinked to the cluster page, or to the chain page for nucleic acid nodes: Clicking each node will display a link below the graph, and clicking the link leads to the "Cluster information" page, and clicking each edge will show an interacting clusters. This page also displays a list of all the member clusters with a summary of each cluster and nucleic acid chains if any. It also shows a list of representative molecules involved in the network. Please note that some networks may be huge and it would take long time to display them.

The example shown here is NetworkID=LN0001