CerebralWeb rendering speeds are between 4.1 and 7.2 times faster than the Cerebral Desktop Cytoscape plug-in (run via Java Web Start) for small and large networks (see the project website for full benchmarking results). CerebralWeb is compatible with current versions of major browsers (Chrome, Firefox, Safari and Internet Explorer) and includes a range of ‘out-of-the-box’ interactive features to support even a novice developer in getting embedded visualizations up and running quickly. Of the several network visualization packages available in JavaScript (see ) Cytoscape.js is by far the most mature, widely used and amenable to plug-in development. CerebralWeb requires jQuery and is compatible with the current version (v2.3) of Cytoscape.js, an open-source graph library available at ( 6).
Cytoscape js example install#
The configuration needed to install the plug-in has been reduced to a minimum by providing default settings and extensive code examples that can be reused and adapted if needed.
The CerebralWeb plug-in is designed to be easily integrated into any website. More information about the web service can be found on the project website and demo pages. The web service query will then look-up the gene ontology (GO) ( 5) Cellular Compartment terms associated with each gene and, using a manually curated map of GO to Cerebral Localization terms, will map the Gene IDs to Cerebral localizations. To access this feature of CerebralWeb the user need only provide Ensembl, Entrez Gene, UniProt or InnateDB identifiers as node attributes. This feature also represents an advance over the Cerebral Desktop application, which requires such annotations to be pre-defined within user supplied network files. Alternatively, CerebralWeb may fetch this information in real-time from ( 3) via a web service that provides Cerebral subcellular localization annotations for human, mouse and bovine genes. The user may specify subcellular localization via the ‘localization’ node attribute in an uploaded network file.
Cytoscape js example series#
The slow execution time of Cerebral visualizations via the Java Web Start also hampers interactivity, especially where a user may wish to perform a rapid series of repeated queries and visualizations, such as when searching a molecular interaction database.ĬerebralWeb can be used to visualize any large network (up to ∼2000 nodes for optimal performance given current JavaScript engines and hardware speeds) where the nodes have an attribute value matching one of Cerebral’s subcellular localizations. CerebralWeb does not require the installation of Java and is not prone to Java related security and accessibility issues that may prevent the execution of Java Web Start dependent applications. CerebralWeb can also interact with other components within the host webpage such as a tabular view of results, a search box or real-time filtering features.
This improves usability for desktop browsers but also adds support for increasingly prevalent mobile browsers. First, the embedded nature and speed of CerebralWeb allows it to be seamlessly integrated into interactive websites. To address these limitations we have created CerebralWeb, a light-weight, web-embedded, JavaScript implementation of the Cerebral concept, which is intended to support network visualization on systems biology and molecular interaction database websites.ĬerebralWeb has a number of advantages over the Java Web Start approach of incorporating Cerebral into web applications. Currently, however, databases that utilize Cerebral, such as ( 3, 4), must do so via a Java Web Start version of Cytoscape, which is slow to load, often subject to accessibility issues and generally prohibits the fast and fluid interactivity required by web applications. Cerebral (‘Cell Region-Based Rendering and Layout’) is an open-source plug-in for Cytoscape that uses subcellular localization information to layout networks in a more biologically intuitive and layered fashion ( 2). In particular, many biologists think of pathways or networks as having a relatively linear underlying structure with a top (cell surface), middle (cytosol) and bottom (nucleus). A critical challenge for the visualization of biological networks is to layout the network in a manner more familiar to a biologist. Network visualization tools, such as the popular Cytoscape application ( 1), allow researchers to interactively visualize these networks and to discern biologically meaningful patterns that may otherwise be difficult to detect. A key goal of systems biology is to understand the emergent properties of molecular interaction networks and to investigate how these properties are dysregulated in disease.
0 kommentar(er)
