Since August 2008, the eScience
Research Group in the Department of
Computer Science at the University of
Virginia has been participating in the cancer
Biomedical Informatics Grid (caBIG) Architecture Workspace .NET Working Group to create a design
to support caBIG and caGrid
using .NET. More specifically, we have prototyped .NET-based clients and
services that show proof-of-concept for future .NET-based caGrid
services. This project produces a comprehensive plan for how to utilize .NET in
the future of caGrid. Clearly, a .NET-based
implementation of caGrid would greatly enhance the
options available to caGrid developers and deployers.
Please contact me (Marty Humphrey) for more
information!
[update
– November 6, 2010] We just
attended the caBIG Tissue Banks and Pathology Tools Face-to-Face
Meeting in Houston (Nov 2-4). While there, we were asked by Dr. Ian Fore, NCI
Coordinator for the workspace, if we could create functionality that would
allow a Microsoft Excel user to create Global Specimen Identifiers while still
using Excel. It’s still a prototype and needs to be cleaned up a bit, but
here’s a 2-minute video of it in action…..
[update – Oct 9, 2010] We led a session at the
hack-a-thon at the 2010 caBIG Annual Meeting! We
created a lab based on the use of Windows Azure
for caBIG. We will put those materials and a
screencast of that up shortly. In the meantime, here’s a screencast of
our most recent development work for caBIG using .NET
(these materials will also be up shortly).
In the following screencast, the
scenario is that an enterprise has a database of Tissue Samples, and they want
to make the existence of these samples known to the caBIG
cancer research community. So they want to participate as a service in the Common Biorepository Model (CBM) effort in caBIG.
So the problem they are facing is: how can they best leverage
their existing expertise with .NET, Microsoft SQL Server, Visual Studio, etc.,
and can they use their existing DB (because it basically contains the same
information as in the CBM effort, albeit in a slightly different schema)?
In this screencast, we show how
to use Visual Studio 2010 Ultimate with the Visualization and
Modeling Feature Pack, along with some code that our research group at the
University of Virginia has written, to:
1.Create a new
“modeling project” in Visual Studio and import the CBM UML (XMI)
2.Map the UML
model onto the existing SQL Server 2008 DB (without changing the schema of the
DB)
3.Auto-generate
the .NET service “guts” from this UML model (resulting in a .NET
CBM DLL tailored to the schema of the back-end DB)
4.Create a Windows
Azure service (note: in the screencast,
we just use the Windows Azure development fabric, but we could have used
Windows Azure itself)
a.We first store
this DLL and a “configuration” file in Windows Azure storage
b.We then
configure our Windows Azure service to read its “guts” from Windows
Azure storage (the DLL and configuration file from the previous step)
c.We then deploy
the service
We then run a “tissue
sample searcher” application to show that the new .NET-based service
answers CBM queries (e.g., “Which
Prostate Gland samples do you have?”)