ConsensusPathDB

Category Cross-Omics>Knowledge Bases/Databases/Tools

Abstract ConsensusPathDB is an interaction meta-database that integrates functional interaction repositories forming a heterogeneous ‘interaction network’ which comprises protein-protein interactions, as well as signaling, metabolic and gene regulatory interactions.

ConsensusPathDB integrates different types of functional interactions between physical entities in the cell like genes, RNA, proteins, protein complexes and metabolites in order to assemble a more complete and a less biased picture of cellular biology.

Currently, ConsensusPathDB contains metabolic and signaling reactions, physical protein interactions and gene regulatory interactions in human, mouse and yeast.

Physical entities from the external resources are mapped to each other on the basis of common identifiers like UniProt and Entrez.

Interactions with matching primary participants (i.e., substrates and products in the case of biochemical reactions, interactors in the case of protein interactions and regulated gene in the case of gene regulatory interactions) are also mapped and grouped together according to similarity.

The database content is updated every three (3) months with the newest available versions of the source databases.

Data currently (December 2011) originates from 24+ public resources for functional interactions (see below...), as well as interactions that the manufacturers have curated from literature.

Additionally, biochemical pathways have been imported from several databases for use in pathway analyses. Data are integrated in a complementary manner and redundancies are avoided.

ConsensusPathDB - Human Integrated databases --

Reactome - (see G6G Abstract Number 20267); Kegg; Humancyc - (see G6G Abstract Number 20234);

Pathway Interaction Database (PID) - (see G6G Abstract Number 20245);

Biocarta; Netpath; Integrating Network Objects with Hierarchies (Inoh) Pathway Database; Ehmn; Intact-ss (molecular interaction database); Intact-ls; Database of Interacting Proteins (Dip); Molecular INTeraction database (MINT);

Human Protein Reference Database (HPRD); Comprehensive Resource of Mammalian protein complexes (CORUM); Biological General Repository for Interaction Datasets (BioGRID);

MIPS Mammalian Protein-Protein Interaction Database (Mips-mppi); BIND; MatrixDB; SPIKE (Signaling Pathway Integrated Knowledge Engine) - (see G6G Abstract Number 20127);

Pathogen Interaction Gateway (Pig); PhosphoPOINT; PDZBase; InnateDB; PhosphoSitePlus; Pharmgkb - (see G6G Abstract Number 20304);

The Small Molecule Pathway Database (SMPDB); SignaLink (a signaling pathway resource); and WikiPathways - (see G6G Abstract Number 20367);

ConsensusPathDB statistics --

Physical entities: 52,406

Interactions: 161,602

Pathways: 3,186

ConsensusPathDB Interactions of specific molecules or pathways --

The ConsensusPathDB user can search for interactions of specific physical entities or pathways.

First, search terms are provided in the form of trivial names or accession numbers (e.g., UniProt or KEGG accession numbers). Search results summarize the names of matching objects and, if the objects have been searched by accession numbers, the accession numbers matching the query.

Additionally, external web links are provided for each hit that shows the origin of the physical entity or pathway, and navigates the user to the web-site containing the original information about the item.

After selecting relevant pathways or physical entities, their functional interactions are listed. If the selected pathways contain sub-pathways or if the physical entities constitute groups, e.g. protein families, then the interactions of the sub-pathways or group members, respectively, are also listed.

At this stage, each interaction that is displayed has a single source. The user is able to specify mapping criteria according to which similar interactions are to be merged. Interactions with matching primary participants are considered similar.

Which interactions should be considered identical, depends on the user's settings: the user is able to specify whether the primary participants of similar interactions must have matching modification patterns, subcellular localization and/or matching stoichiometry (in the case of biochemical reactions) in order for the similar interactions to be considered identical.

According to these settings, interactions are merged together and displayed in the visualization environment of ConsensusPathDB.

Shortest path of functional interactions between molecules --

Apart from the standard search for interactions of physical entities and pathways, the web interface of ConsensusPathDB features searching for the shortest path(s) of functional interactions that link a couple of physical entities with each other.

Here, the user specifies a path “start” and a path “end”. One possible shortest path between both is calculated and displayed.

The user is able to exclude particular physical entities from the shortest path, which is especially useful when non-specific hubs like ubiquitin and ATP are present in the path. Interaction paths of interest can be visualized in the visualization environment of ConsensusPathDB.

ConsensusPathDB Interaction network visualization environment --

Interaction networks can be viewed in an interactive visualization environment. The web interface user has two (2) choices for a visualization environment: an image/JavaScript-based visualization framework and a Java applet-based framework.

Both frameworks display interaction networks in the same style so switching between them involves No user acclimatization.

In the Java-based visualization environment, gene / protein expression data can be overlaid on the nodes of a currently viewed network to enable the interaction network based interpretation of such data.

ConsensusPathDB Interaction network miscellaneous functions --

For interaction networks displayed in the visualization interface, the user can view a summary of the number of physical entity and interaction nodes of specific types, as well as an interaction mapping summary. Visualization graphs can be exported from the visualization environment.

Several possibilities are provided: the interaction network-graph can be exported as a graphics file in several formats, including png, jpg, postscript and others, as a BioPAX file or as a ConsensusPathDB model dump.

BioPAX is an XML-based file format that carries information on interaction systems and is currently supported by many software packages for e.g. biochemical system modeling.

The ConsensusPathDB model dump can be imported into the visualization environment and worked with at a later time point.

ConsensusPathDB Pathway analysis --

ConsensusPathDB offers two (2) statistical approaches to analyze user-specified lists of genes obtained, e.g., by microarray experiments (commonly, a list of differentially expressed genes between two phenotypes).

The first approach is over-representation analysis, where predefined lists of functionally associated genes [pathways, Gene Ontology (GO) categories and neighborhood-based entity sets] are tested for over-representation in the user-specified list based on the hypergeometric test.

The second approach is based on the Wilcoxon signed-rank test and takes as input, genes with exactly two (2) measurement values, typically expression values in two (2) distinct phenotypes.

While the over-representation functionality typically takes as input a relatively short, non-weighted list of “special” (e.g., differentially expressed) genes, for the Wilcoxon enrichment analysis approach, genome-wide expression data containing measurements of possibly thousands of genes are the preferred input.

ConsensusPathDB Network upload --

Through the web interface of ConsensusPathDB, users can upload interaction networks in BioPAX, PSI-MI or SBML format. Upon upload, physical entities and functional interactions are compared against the content of ConsensusPathDB. This is done based on identifiers (UniProt, KEGG, ChEBI ...) in the case of physical entities and on participant composition in the case of interactions.

The uploaded networks are visualized, and for interactions present in ConsensusPathDB, the source databases are displayed. Networks can be expanded in the context of the database content by adding further interactions, and editing.

Note: SBML files should have physical entity annotations as in the BioModels model repository - (see G6G Abstract Number 20295).

ConsensusPathDB plug-in for Cytoscape --

With the ‘ConsensusPathDB plug-in for Cytoscape’ (version 1.0) you can mine evidences (publications, detection methods, pathways, etc.) for interactions loaded into Cytoscape - (see G6G Abstract Number 20092) and highlight ‘interactions’ that have Not been detected previously.

System Requirements

We-based.

Manufacturer

Manufacturer Web Site ConsensusPathDB

Price Contact manufacturer.

G6G Abstract Number 20690

G6G Manufacturer Number 104265