genSpace Project

Many researchers in the field of computational biology use the geWorkbench tool, a Java-based open-source platform for integrated genomics. Using a component architecture it allows individually developed plug-ins to be configured into complex bioinformatic applications. At present there are more than 30 available plug-ins supporting the visualization and analysis of gene expression and sequence data, however it is purely a stand-alone application and does not have any collaboration capabilities.

This project, started in Fall 2007, is designed to develop proof-of-concept collaboration facilities suitable for inclusion in geWorkbench and, at least in principle, other plugin-based tool integration environments. The project aims to include a few of the well-known collaboration mechanisms, possibly including but not necessarily limited to instant messaging (as in AIM), social networking (as in Facebook), knowledge repositories (as in Usenet newsgroups) and shared folders (as in Webdav). However, only those existing collaboration mechanisms that will in some manner facilitate new research towards novel mechanisms will be chosen.

Some basic functionality has already been implemented: logging which analysis tools users run, the ability to log in to genSpace, rudimentary integration with the jClaim chat client, and a server-side instant messaging bot that allows for simple user interaction, such as the creation of social networks.

In Spring 2008, we integrated all of this functionality into an end-to-end system which allows all geWorkbench users to log into genSpace, communicate with other users, find experts on the tools they are using, and find other researchers who are similar to them. We also built visualization tools that allow users to see their social networks, and to see "collaboratively created workflows", which are automatically generated based on other users' activities.

This summer, we are continuing to develop and refine those features, and we will begin data collection through the logging of real geWorkbench users' activities. We are also planning on implementing a recommendation system, in which genSpace can recommend actions based on what has been observed.

Faculty
  Prof. Gail Kaiser, kaiser [at] cs.columbia.edu

Graduate Students
  Chris Murphy, cmurphy [at] cs.columbia.edu
  Swapneel Sheth, swapneel [at] cs.columbia.edu
  Cheng Niu, cn2198 [at] columbia.edu

Former members
  Diana Chang, dc2199 [at] columbia.edu