Accelerating Research with Effective Biobanking
caBIG™ in action at Washington University School of Medicine’s Alvin J. Siteman Cancer Center
There was a time not too long ago when accessing tissue samples from the biobank to conduct research at the Siteman Cancer Center involved a plethora of different people and multiple manual steps. Typically, a researcher would request a specific specimen by sending an email to the biobank manager, who manually sifted through charts and tables and then returned an email response to the researcher, usually with additional questions, before appropriate specimens were found. Researchers were frustrated with the lack of a user-friendly system that they could search directly.
Now, with the development and adoption of caTissue Core, developed within the caBIG™ program under the leadership of Washington University School of Medicine, researchers can quickly and efficiently find the biospecimens they need—reducing the time required from a few weeks to a few days. Even more importantly, having access to more specimens that meet specific study criteria has resulted in more robust and reliable research.
While researchers from Indiana University, Thomas Jefferson University, University of Pittsburgh Medical Center, and the University of Pennsylvania helped develop the system requirements, specifications, and the documentation, and conducted all usability testing, the code development of caTissue Core was led by Nagarajan’s group at Washington University. (Learn more about caTissue implementation at these institutions).
"Ultimately, the passion of the medical community is to make the standard of care better," explains Rakesh Nagarajan, M.D., Ph.D., Co-Director of the Bioinformatics Core at the Siteman Cancer Center at the Washington University School of Medicine.
"By facilitating data storage and analysis, we have made a significant difference and are able to streamline the whole research process. I firmly believe that interoperable tools will improve how diseases are diagnosed at early stages, how they are managed, and, eventually, how diseases are cured."
Building an Information Network: Development and Adoption of caTissue
The Nagarajan laboratory has been instrumental in the development of these caBIG™ tools:
- caTissue Core: A tissue bank repository tool for biospecimen inventory, tracking, and basic annotation.
- cancer Function Express (caFE): A tool that annotates microarray probes using publicly available biomedical databases.
- GeneConnect: A genomic identifier mapping service in which one or more commonly used genomic identifiers is linked with identifiers from other databases.
- caBench-to-Bedside (caB2B): A tool that permits researchers to query caGrid data services. For example, microarray data from caArray can be obtained and then associated with biospecimen information from caTissue.
Nagarajan’s group worked with Mark Watson, M.D., Ph.D., Director of the Siteman Cancer Center Tissue Procurement Facility, to develop caTissue Core, a biobanking management tool designed to collect, manage, process, annotate, request, and distribute biospecimens and associated information.
"The one thing that is really accelerating research and our ability to leverage all of these biospecimens is putting them into an information network where scientists, as a group, can know what’s available for research purposes," Watson notes.
"caBIG™ tools are enabling research progress at Washington University in a fundamental way," adds Nagarajan. "Previously, the tissue bank was the bottleneck. That problem is now greatly reduced by users querying the interface directly, and research progress is significantly accelerated."
Nagarajan’s programming experience and Watson’s biobanking expertise were essential for effectively loading all the data from the Siteman Cancer Center Tissue Procurement Facility into caTissue. This project allows researchers from anywhere on the Washington University campus to use caTissue to easily and quickly access biospecimens and associated data for their own research projects.
During bi-weekly sessions, Watson works with smaller biobanking laboratories on the Washington University campus to map and migrate their data into caTissue. These sessions serve both to educate potential adopters about the management capabilities of caTissue and to demonstrate to groups that have "homegrown" repositories that caTissue will add value to existing systems that depend on lab-specific software. Watson also reassures groups that caTissue will properly manage biospecimen access, annotation, and delivery needs.
Although Nagarajan and Watson are the only instructors at Washington University currently qualified to teach caTissue to end-users, their training sessions include informatics specialists from the Bernard Becker Medical Library who plan to add caBIG™ end-user guidance to the library’s existing training programs.
Sharing the Vision of an Interconnected Research Community
Looking back on his involvement in the project, Nagarajan offers the following tips to newcomers interested in getting connected with caBIG™:
- Take at least 1-3 months to review existing caBIG™ resources like the Tools Inventory.
- Find and contact the appropriate workspace leads, developers, and contact persons.
- Post questions on the developers’ forums on the GForge website—all members of the group dedicated to that specific resource can offer suggestions.
- Learn more about the Cancer Data Standards Repository (caDSR), the NCICB Resource for Common Data Elements.
- Try online or live caBIG™ training.
At Washington University, there is a history of senior leadership support of caBIG™ efforts. Early on, Jeffrey Milbrandt, M.D., Ph.D., a well-respected Washington University researcher, inspired senior leadership with the idea of a bioinformatics core and the vision of tools to interconnect the research community.
Nagarajan continues to keep senior leadership well informed about the caBIG™ project and manages expectations regarding the required time to develop and adopt tools. He and his team are clear about what tools can or cannot do, explaining that a tool "may not be perfect, but may be the best tool available."
"Tissue banking using caTissue will permit a fundamental change in cancer research," says Nagarajan. "There will be more robust findings with larger datasets, especially in studies of rare tumors. Further, caBIG™ will be a model for data sharing in non-cancer disease areas as well."
To advance the concept of an interconnected research community, Nagarajan also draws upon a team of people with different strengths to more effectively manage caBIG™ development projects. "I have found that one person is not enough; a group of five to six people working at 10 to 15 percent is most effective," he says.
Nagarajan recommends that a caBIG™ implementation and adoption team include at least one of the following: a medical informaticist with knowledge of medicine, electronic health records, and clinical trials; a bioinformaticist with knowledge of microarrays, proteomics, and biological research techniques; and programmers and technical people with knowledge of UML and Grid/Web services.
In addition to his team of local experts, Nagarajan also employs the services of an offsite programming team in India (Persistent Systems). Nagarajan notes that the company has been critical in developing, supporting, and maintaining caTissue.
Washington University also has two active nodes on caGrid, so that those outside of the campus can access both GeneConnect and cancer Function Express (caFE). More nodes will be activated when caArray 2.0 and caTissue Suite, the comprehensive version of caTissue, are released.
"caBIG™ infrastructure is fundamental for inter- and intra-institutional data sharing," explains Nagarajan. "Local vocabularies are fine for local projects, but they are neither scalable nor standardized. However, the caBIG™ infrastructure formalizes metadata specifications, which is especially important when trying to semantically interoperate across many nodes."
Learn More:
