Even though this blog was initially created to cover our BioGPS gene portal application, one of the broader themes of our group’s research is harnessing community intelligence in biology. And aside from BioGPS, our other main project in this space is the Gene Wiki initiative.

I’m pleased to report that the NIH (and specifically NIGMS) has decided to support our Gene Wiki development efforts for the next four years. Since the success of the effort really depends on participation by the community, it only makes sense to be completely open about the grant’s aims.

Broadly speaking, our future work will fall under three aims:

  • Ensure structured content is up to date. As the volume of gene annotations in public databases increases, the Gene Wiki infoboxes should be updated to reflect that. In addition, users should be able to quickly and easily create new Gene Wiki pages if a page for their gene of interest does not already exist.
  • Improve reliability, and provide reliability metrics. Wikipedia’s model doesn’t allow readers to evaluate contributions based on the editor’s credentials. Therefore, we are partnering with the WikiTrust team to incorporate content-driven trust metrics. In short, authors earn trustworthiness based on how long their previous edits persist, and each word is assigned a trustworthiness score based on the authors who were responsible for adding and preserving it. Be sure to try the prototype Firefox plugin.
  • Make content computable. Wikipedia is primarily composed of free text, images, and diagrams. This unstructured format is great for reading by humans, but quite difficult for bioinformatics and statistical analyses by computers. As such, a variety of efforts (notably DBpedia) are working on techniques for extracting and distributing structured versions of the knowledge accumulating in Wikipedia. We will be joining this movement towards a more structured, more useful Wikipedia as we explore a variety of strategies for mining, verifying, and sharing new gene annotations from the community-generated content of the Gene Wiki.

More details of our plans and progress in future blog posts. It’s also the right time to introduce Ben Good, a newly-joined group member who will be working on the Gene Wiki full time. Check out his blog, where he often pontificates on various aspects of his projects.

As a community intelligence project, we of course welcome feedback and thoughts. And of course, we most welcome your participation!