Workstream C: Tools

Infrastructure Tools Project
Leads: Carol Goble, University of Manchester, Ted Liefeld, Broad Institute, Alex Pico, Wikipathways/UCSF
Participants: Marc Hadfield, Jim Davies, James Brenton, Norman Paton, Jessie Tenenbaum

Introduction
We plan to design, build and demonstrate an effective infrastructure of support tools for Sage Commons. These tools will support data management, user/group accounts, and research use cases. Core properties of the infrastructure are modularity and web services (or web APIs) that will promote use and reuse of Sage functionality in custom workflows implemented in third-party tools, and to facilitate open-ended innovation in the use of the Sage Commons. With these tools users will be able to access, query, analyze, visualize, annotate, manage and share the network models produced and distributed by Sage Commons.

Activities before Congress
We will adopt a set of existing tools (Alitora Systems, Cytoscape, GenePattern, PathGrid and Taverna) to manage networks, perform Key Driver Analysis, visualize results and annotate and share findings with colleagues. These tools will be modified or extended as needed to interoperate in a loosely-coupled manner through web and web-services interfaces.  Specific tasks include:

  • Load example Sage networks into Alitora Systems, providing user/group support, semantic database of networked data, and web API access to networks. In the process, we will define an initial semantic data model (in coordination with other Sage Data Format experts) to facilitate data flow across the tools.
  • Transform Key Driver Analysis (KDA) into a GenePattern module, providing a web API interface (SOAP) and graphical user interface that expose the analysis method to a variety of applications. The KDA module will be made available for download and will also be made available for execution on a public GenePattern server.
  • Customize KDA module to produce an output file in the Cytoscape format, allowing visualization of the input network and key driver results.
  • Build a GenePattern workflow (pipeline) demonstrating the use of the KDA module from retrieval of a network from the Sage website through the launch of Cytoscape to view the KDA results. This task demonstrates the potential for integration of Sage data and KDA with the other 120+ analyses available within the GenePattern environment.
  • Transform KDA into a PathGrid module, providing another example of a web API interface (SOAP) available to internal and external tools. Build a PathGrid workflow, demonstrating interoperability with other PathGrid modules.
  • Build a Taverna workflow that ties 1-5 all together. A single workflow could automate the process of loading a Sage network via Alitora, performing KDA (via GenePattern or PathGrid), and visualizing the results in Cytoscape. This task immediately opens up the possibility of integrating other Taverna services into the workflow per the user’s interest and expertise. This task also demonstrates the openness of the proposed approach by demonstrating a third workflow and analysis environment capable of operating on the Sage Commons.
  • Within Cytoscape, the Alitora plugin can support extending/augmenting the network using annotations, ontologies or other networks. The network or select components can be saved to the user’s account and then further annotated and shared with colleagues through Alitora’s web interface.

Activities at the Congress

We will have a working demonstration of the infrastructure described above and prepare a screenshot slideshow.   This will be presented as a draft set of tools that can provide an analysis infrastructure for the Sage Commons. We will highlight the strengths of what we’ve envisioned, focus on the core principles behind the approach and will solicit feedback from the assembled community for alternative or additional tools. We will collect as many specific use cases from the community to specifically address feature prioritization. We will summarize the experience of putting this trial infrastructure together (i.e., lessons learned) and discuss the implications for the Sage Commons requirements and specifications.