NCI Cancer Bulletin: A Trusted Source for Cancer Research NewsNCI Cancer Bulletin: A Trusted Source for Cancer Research News
March 2, 2004 • Volume 1 / Number 9 E-Mail This Document  |  Download PDF  |  Bulletin Archive/Search  |  Subscribe

The information and links on this page are no longer being updated and are provided for reference purposes only.

Special ReportSpecial Report

caBIG: The Launch of a Bioinformatics Community

When NCI announced the pilot of the cancer Biomedical Informatics Grid, or caBIG, last summer, it marked the beginning of a potentially powerful avenue for cancer research that is open to the entire cancer community. With caBIG, all cancer researchers will have access to a common research infrastructure that creates a plethora of opportunities to not only make important new findings but to do so more quickly and efficiently than ever before.

Currently in its early stages of development, this new system will offer a library of tools and resources - from clinical trial management systems to tissue bank and pathology tools - that are all built to common standards and are interoperable with other existing systems. It will also allow researchers to tap into an ocean of raw published data. As currently envisioned, a researcher could, for example, use caBIG to tap into appropriately anonymized, published molecular data from all U.S. cancer centers on patients with prostate cancer who are being treated with a specific drug. By providing such access supported by these innovative tools, study populations' data can be far more robust and researchers can mine the data in a way that simply isn't possible at the moment.

The Launch of a Bioinformatics Community Technically, caBIG will work much like a home entertainment component system in which components are made by different manufacturers. Each component - the CD player, the FM receiver, the speakers, etc. - is built to certain common standards. As a result, each component, no matter who manufactures it, can be hooked up to the other and they will all work seamlessly. To date, this is not the case with applications and data generated within the cancer research establishment.

"Each cancer community has its own dialect, its own standards for capturing and reporting data, and each cancer center builds its own unique software, its own clinical trial management systems, its own gene expression array data capture systems," says Dr. Ken Buetow, the project director for caBIG. "While in and of itself this is not bad, there is no interoperability between systems from cancer center to cancer center - or commonly within a single cancer center - meaning that volumes of valuable raw data are not being tapped, effective best practices are not being widely distributed, and resources are being wasted because of duplication of effort.

"An increasing fraction of a scientist's effort is spent managing and manipulating the ever-growing, complex datasets being generated by modern biomedical studies," he continues. "Institutions are spending vast resources reinventing basic infrastructure to support these activities. Yet the creativity in being a scientist, and the impact of our research, is really about answering questions. With caBIG, the research community can focus its attention on innovation in prevention, diagnosis, and treatment, not on data management and constantly building and troubleshooting the basic underlying research infrastructure."

The caBIG initiative represents NCI's response to calls from the community to assist it in dealing with the tsunami of biomedical research data, and its development has been a community-wide effort, involving more than 50 cancer centers.

The Pilot
The central element of the initial three-year caBIG pilot is represented by the "workspaces" - virtual environments where related caBIG resources and tools are grouped. In individual workspaces, staff from participating cancer centers will work on projects as both "developers," that is, those that develop tools and resources based primarily on existing and proven items, and "adopters," those that test, validate, and apply the tools and who contribute data resources used in this validation. There are also working groups that provide guidance and support to each workspace and the pilot as a whole.

These are truly collaborative efforts. The clinical trial management systems workspace and working group, for instance, include 15 participating cancer centers. The integrative cancer research workspace and working group have 23.

As Dr. Buetow points out, the intention of caBIG is not to reinvent the whole system of cancer research. In each strategic area of the pilot, existing tools and resources are being assembled whenever possible and adapted to become compatible with caBIG. The projects being tackled in this three-year pilot are those that were identified by the cancer centers as representing the greatest areas of need.

"We couldn't do everything at one time," says Dr. Buetow. "A lot of worthwhile projects were suggested. Working with the cancer centers, we are focusing on tools and resources that will be most immediately helpful. We will definitely get to the other projects over time - developing, testing, and validating them as needed, based on input from the community."

The goal for caBIG, Dr. Buetow stresses, is that it will ultimately evolve into a community-driven activity, with the effort expanding beyond the groups directly supported by NCI, including industry.

"If caBIG meets its pilot objectives, its value to the cancer research community will become clear beyond any 'measurements' we can make," Dr. Buetow says. "We clearly have high expectations for caBIG and, based on the initial response we've had and the progress we've made to date, I'm confident that we'll meet and possibly surpass them."

More information on caBIG, including a virtual tour, is available on the NCI Web site at