The Biopathways Workbench design provides a suite of applications and a programming toolkit to help create, edit, model, and visualize signaling and metabolic pathway data. Workbench applications will import and export data in a variety of file formats, access on-line databases, and visualize large-scale networks in 2D and 3D.
The
central organizing metaphor for Workbench data is a reaction network graph, or pathway,
with chemical compounds represented as nodes, and chemical interactions
represented as links, or edges, between the nodes.
Numeric, text, and structured data attributes will be bound to nodes and edges, including molecule names, synonyms, locations, concentrations, charges, interaction types, molecule roles in those interactions, and so forth. Hierarchical classification schemes, such as those from the Gene Ontology database, provide additional annotations that characterize types of nodes and edges. Together these data will be used to guide data filtering and visualization.
Learn more about the pathway data model design
A pathway can be built as a single large collection of chemical compounds and interactions, or as a hierarchy of pathway modules. Each module encapsulates a subset of chemical compounds and interactions, presenting them as a single icon that can be "opened" to reveal the module's detail. Modules may contain other modules, enabling an author to express a very complex pathway as a hierarchy of pathways within pathways showing progressively finer detail as the user drills deeper into the pathway.
Embedded domain knowledge will enable the Workbench to intelligently assist in creating and simulating valid pathways. Chemical data may be looked up in on-line databases of proteins, small molecules, interactions, and pathway modules.
Interactive
visualization supports direct manipulation of pathway components in a
click-and-drag style similar to drawing tools. Chemical compounds may be added,
deleted, and moved, and interactions drawn between them. Individual components or groups of components may be selected, cut, copied, and pasted. A double-click atop a component shows its editable attributes, including its name, synonyms, type, description, database IDs, time series data, and simulation parameters.
3D visualization arranges nodes in depth as well as side to side and top to bottom. Related pathways may be layered and aligned to highlight simularities and differences. Numeric data may "hover" near molecules and interactions, while animations show the flow of a signal through a simulated pathway.
Learn more about pathway visualization design
The Workbench architecture design is centered on a set of core services supporting data structure management, model building, and visualization. Plug-ins attach to these services to provide support for specific file formats and databases, and to map between the native data structures of these data sources and those of the Workbench. Similarly, plug-ins may analyze the data, validate it, filter it, alter it, amend it, or derive new data. Visualization plug-ins may perform layout functions or assign visual attributes based upon the data's own attributes and annotations.
The Workbench is file format neutral, importing and exporting data in a variety of file formats, including its own PATH file format. The Workbench will support standards in the field, such as KEGG's KGML, SBML, and BioPAX. Additional file formats may be supported via plug-ins.
Learn more about file formats design
Plug-ins provide access to on-line databases, such as the LIPID MAPS pathway and compound database. Additional databases may be supported via plug-ins.
Learn more about database design
The Workbench will be available for computers running Windows XP (or later), Mac OSX 10 (or later), and all variants of Linux for which Sun's Java and Java OpenGL (JOGL) are available. Java's OpenGL enables Workbench applications to use the 3D graphics abilities of today's video cards to speed up visualization tasks.