Pathway Explorer (Figure 4) provides a generic user interface for posing a large class of queries about pathways using tree-structured views. Entities involved in the pathways database can be better viewed by generalization/specialization hierarchies, e.g., classes of pathways, super pathways, functional classifications in arbitrary granularities, etc.. In addition, pathway components, i.e., processes in a pathway, and molecular entities involved in each process can naturally be viewed in the form of a tree-structured (hierarchical) view. Motivated by this observation, Pathways Explorer is a query interface which allows users to query the database using tree-structured views. As an example, consider the tree-structured (hierarchical) schema Pathways-Process-Molecular entity, where a pathway consists of several processes, and each process is associated with several molecular entities. This may also be visualized as a logical schema of a nested relation (Pathways (Process (Molecular_Entity)*)*), or an XML view. Thus, Pathways Explorer can also be considered as a query interface for querying tree-structured views for a relational database.
Given the schema of an XML view, or a nested relation view of the underlying relational database, query interface (template) structure follows that of the view schema. The schema declares that pathway elements contain process elements that contain molecule elements, so the tree in the user-interface would contain only these nodes, along with the node attributes as children. Each attribute’s possible values may then be seen individually by opening its drop-down box. Choosing a value for an attribute effectively creates a query constraint declaring that only nodes with the chosen attribute value are of interest. For more details on the Pathway Explorer, see Pathway Explorer user manual.
Figure 4. A snapshot of Pathway Explorer.
Pathway Editor allows view-based database updates; that is, users can insert/update/edit pathways via the high-level user views of the pathway viewer. The pathway editor, while enforcing the integrity of the existing database (sometimes interactively with user input), maps them as updates to the underlying relational database. This component is currently functional for metabolic pathways. Pathway Editor is initiated from context menu of the tree node in Pathway Browser. To learn more about the Pathway editor, see Pathway Editor user manual.
Figure 5. Snapshot to launch Pathway Editor.
The Pathway web services component provides many querying functions that are available over the Internet. This component has a function for each query it implements. Pathway tools in any machine connected to the Internet can use SOAP (Simple Object Access Protocol) to invoke these functions, and receive query results. We have implemented over forty functions. Querying functions are categorized into five types: molecular entity queries, reaction (process) queries, pathway queries, organism queries and path computations. These queries are illustrated in Figure 6, and can be invoked from the website http://nashua.cwru.edu/pathways_metabolic_1/service1.asmx
Figure 6. Pathway web services.
