---HOME--- Browse the database Interactome Construction Statistics Download Tutorial - Help
MatrixDB banner

tutorial_iconMatrixDB Tutorial

Last updated: 2010/07/29

MatrixDB Content | Searching MatrixDB | Biomolecule data | Interaction data | Download | Interactome Construction | Publication

MatrixDB content:

Organization of MatrixDB database:

m spacer sprot spacer CheBI


spacer HPRD space dip space intact space mint space biogrid


flecheBack to the Top

Searching MatrixDB:

0

Display all the molecules of a category

Display all the molecules corresponding to the selected category. "All extracellular matrix biomolecules" shows all the molecules annotated by the UniProtKB/Swiss-Prot (version 2010_06) keyword "Extracellular Matrix" and by Gene Ontology terms corresponding to extracellular matrix.

Searching for specific biomolecules

The search can be performed using Results are displayed in a table. Moving the mouse over the name displays the full name of the molecule. Click on the molecule name to access data (Biomolecule data).

Advanced searches:

1

Interaction data from specific sources

Displays the interactions of a particular dataset: interactions imported from databases (IntAct, MINT, DIP, BioGRID and HPRD), from literature curation (MatrixDB Literature curation), from experimental data (SPR arrays data) or specific interaction datasets (e.g. elastic fibers interactome)

Free text search

The free text search queries a large part of the fields stored in MatrixDB. A search by keyword (collagen, basement membrane, Alzheimer disease... ) or by author name can for example be used in this search box.

PubMed IDentifier (PMID) search

PubMed IDentifier (PMID) is the publication identifier in PubMed (e.g. PMID: 19147664). The PMID search can be used to check if the interaction data reported in the publication are stored in MatrixDB.

IMEx IDentifier (IMEx-ID) search

The IMEx identifiers are common identifiers for publications exchanged within IMEx consortium, allocated by IMEx Central (IMEX-x) and for experiments (IMEX-x-y). The IMEx search can be used to access directly to specific publication or experiment using IMEx identifiers.


flecheBack to the Top

Biomolecule data:

Protein data

Proteins are identified by their UniProtKB/Swiss-Prot primary Accession Number (Example: P98160). This is the more stable identifier of UniProtKB/Swiss-Prot (when a protein entry is modified, the previous accession number is retained in a secondary accession number list).
flecheBack to the Top

Glycosaminoglycan data

Glycosaminoglycans are identified a MatrixDB identifier "GAG_x" where x is a number, and cross-referenced to the ChEBI (Chemical Entities of Biological Interest) database.
flecheBack to the Top

Protein Fragment data

Protein fragments are identified by a MatrixDB identifier "PFRAG_x_species" where x is a number and a cross-reference to the PRO feature of UniProtKB when available.
flecheBack to the Top

Cation data

Cations are identified by a MatrixDB identifier: "CAT_x" where x is a number and a cross-reference to the ChEBI (Chemical Entities of Biological Interest) database.
flecheBack to the Top

Lipid data

MatrixDB identifier: "LIP_x" where x is a number and a cross-reference to the ChEBI (Chemical Entities of Biological Interest) database.
flecheBack to the Top

Multimer data

Multimers are identified by a MatrixDB identifier "MULT_x_species" or "MULT_x_VARy_species" for molecular isoforms where x and y are numbers and a cross-reference to EBI complexes when available
flecheBack to the Top

Interaction data:

Interactions are identified by a MatrixDB interaction identifier: identifiers (UniProtKB acession number for proteins and MatrixDB identifiers for other molecules) of the participants, separated by an underscore in alphabetical order.
flecheBack to the Top

Experiment data:

MatrixDB experiment identifier: interaction identifier (= identifiers of the biomolecules in alphabetical order separated by an underscore), followed by the PMID (PubMed Identifier) and the number of the experiment (1 for the first experiment reported in a paper).

HPRD data

MatrixDB and other IMEx partners data (MINT, DIP, IntAct, BioGRID)

Experiments are reported according to the Minimum Information required for reporting a Molecular Interaction experiment (MIMIx) or to the International Molecular Exchange (IMEx) curation rules. As an IMEx member, MatrixDB uses the controlled vocabulary browsable here.
flecheBack to the Top

Download MatrixDB data:

TutoDownload








A click on the Download button gives access to the last release of MatrixDB curated interaction data under PSI-MI XML 2.5 or MITAB 2.5 format.
The HUPO Proteomics Standards Initiative (PSI) defines community standards for data representation in proteomics to facilitate data comparison, exchange and verification. The PSI-MI XML and MITAB formats are data exchange formats for molecular interactions. More information is available on the PSI Website. A permanent download link is also available at the bottom of this page.


flecheBack to the Top

Interactome Construction:

A interaction network or interactome can be constructed directly from the "Interactome Construction" or from the "Biomolecule report" pages. Two possibilities are available on the Biomolecule report: a direct link to the interactome construction (On the left: "Construct the interaction network of this molecule") that fills the biomolecule identifier field or a button that add the biomolecule to the cart (On the right: "Add this molecule to the list") that store molecules to construct their combined interaction network.

4a

When you click in the button "Add this molecule to the list" the message is modified and becomes "Biomolecule added to the interaction netwok". You can access to the cart by clicking on the "Access biomolecule list" link.

4b

On the cart page, the molecules added to the cart can be visualized and removed if necessary. The molecules can then be sent to the "interactome construction" page to construct the combined interaction network of the selected molecules

4c

There are several possibilities to build interaction netwoks: 2
2b

It may take a few minutes to build an interaction network.
Once the interaction network is ready, the following page will be displayed.


3

"Click to show the interaction network using Cytoscape": this launches the Cytoscape application, which builds an interactive graphical representation of the network.
"Click to show the interaction network using Medusa": this launches the Medusa application, which builds an interactive graphical representation of the network on MatrixDB website.
"Click for another query": return to the interactome construction page.
"Cytoscape Information" gives a brief overview of Cytoscape and a link to the official Cytoscape Tutorial.



TutoConstruction3 TutoConstruction4











Upon launching Cytoscape a first message asks which program to use to open the file. This file must be opened with Java Web Start, which is available in the Java Runtime Environment. If this program is not available on your computer, download it here. To run Cytoscape, the execution of the application in the warning panel needs to be accepted (click on run). The certificate issued by the CNRS (french National Scientific Research Institute) is unfortunately not recognized as a trusted source by default.

TutoConstruction5














The interaction network is displayed using the default visual style, with specific colors for each class of molecules: green for proteins, teal for protein fragments, yellow for lipids, magenta for cations, orange for multimers, and purple for glycosaminoglycans.
Dashed lines ended by an arrow: covalent link between two molecules.
Dashed lines ended by a diamond: link between the protein fragment and its parent biomolecule.
Dashed lines ended by a circle: link between a monomer and a multimer.


TutoConstruction6








Several representations can be selected, including the Organic (Layout->yFiles->Organic), or the Force-Directed (Layout->Cytoscape Layouts->Force-Directed Layout->unweighted) layouts. The force-directed style gives the following graphs for the interaction network of perlecan (level 2)
.


TutoConstruction9 TutoConstruction6_b TutoConstruction10_b











Visual attributes can be selected in the left panel, in the VizMapper part, or created de novo. Default and localization visual styles have been developped by MatrixDB. For cell localization, extracellular molecules are colored in: blue, secreted molecules
(found in the extracellular space, but are not deposited in the extracellular matrix) in purple, membrane molecules in green, cytoplasmic molecules in  yellow, and nuclear molecules in red. These localization annotations are from UniProtKB/Swiss-Prot, Gene Ontology, and from MatrixDB. Specific categories have been created for molecules existing in several localizations: magenta: molecules in nucleus + extracellular matrix, orange: nuclear + cytoplasm, light green: membrane + cytoplasm, teal: extracellular matrix + membrane.

TutoConstruction11 TutoConstruction8











On the graphs, molecules are identified by UniProtKB/Swiss-Prot or MatrixDB identifiers. Information on the molecules can be displayed by selecting one or several items: GO term - biological process, UniProtKB/Swiss-Prot Common name,
UniProtKB/Swiss-Prot keyword, localization  (UniProtKB/Swiss-Prot, Gene Ontology and MatrixDB annotations) , type of ligand (protein, glycosaminoglycan, multimer) in the bottom panel. The datasource source (MatrixDB, HPRD, BioGRID, DIP, MINT, IntAct) or the methods demonstrating the interaction are displayed upon clicking in the "Edge Attribute Browser" in the bottom data panel. Further information about the Cytoscape Software is available in the Cytoscape 2.6 tutorial or on the Cytoscape WebSite.
flecheBack to the Top

Medusa is a graph visualization tool. Below is an example of interaction network displayed with Medusa:

3

Shifting-clicking on a node opens a new window with Biomolecule data. Moving the mouse over a molecule displays the name of the molecule. "Layout" button displays the interaction network using a Fruchterman Rheingold algorithm (Spring layout). "Relax" option provides an interactive spring layout. Nodes can be moved or fixed using Ctrl+click on a node.

Publication:

Please cite the following reference for the database:
Emilie Chautard, Lionel Ballut, Nicolas Thierry-Mieg and Sylvie Ricard-Blum.
MatrixDB, a database focused on extracellular protein-protein and protein-carbohydrate interactions.
Bioinformatics 2009 25: 690-691. [link]

flecheBack to the Top

Powered by ACeDB Valid XHTML 1.0 Transitional CSS Valide !