NCI CPTI: Technologies for proteomics
Potential biomarkers are churned out of proteomics laboratories every day, but these candidate proteins and peptides are not being used routinely in clinical tests yet. To translate proteomics research discoveries into clinical tests, several hurdles must be overcome. For example, scientists from various labs often find nonoverlapping sets of biomarkers for the same disease. Reproducibility, therefore, is a major challenge for proteomics. In addition, improvements in data analysis, experimental design, and instrumentation could help the field to progress.
For a few years now, these problems have been on the minds of researchers at the National Cancer Institute (NCI). After receiving extensive input from the proteomics community, NCI established the $104 million Clinical Proteomic Technologies Initiative (CPTI) to encourage scientists to systematically evaluate the methods and instruments they have been using and to develop new approaches and reagent resources. In December 2005 and February 2006, NCI announced requests for applications (RFAs) from researchers to tackle these issues. The application period closed this spring, and awardees were notified last month.
To lay the groundwork for CPTI, NCI program directors established a 2-year pilot program, called the Mouse Proteomic Technologies Initiative, in which researchers used proteomics methods to study mouse models of cancer. Two consortia were formed, and each group investigated different technologies. According to Adam Clark of NCI, who is a program manager for CPTI, one consortium tested a high-throughput shotgun approach, and the other tested gel-based methods. Although the researchers are not yet ready to release their conclusions, their experiences helped shape the direction of CPTI, say officials.
caBIG: Connecting researchers
After consulting with researchers at NCI-designated cancer centers a few years ago, administrators realized that a grid linking researchers across the country with each other and with software and data would be a valuable resource. A 3-year pilot of caBIG began in 2004 with scientists at those centers, but the project has since expanded to include a broad range of research, clinical, and commercial institutions, says Ken Buetow, NCI associate director for bioinformatics and information technology and director of the NCI Center for Bioinformatics. At present, >80 organizations and >800 researchers participate in caBIG, and any interested scientist, institution, or company can still join.
“If we really want to solve a complex problem like cancer, we have to have a means to integrate data from many sources,” says Buetow, and caBIG is one way to do just that. Data and software for many aspects of biomedicine, such as proteomics, clinical trials data management, tissue banks and pathology, and in vivo imaging, are available on the grid. caBIG can be accessed at https://cabig.nci.nih.gov.
The first CPTI RFA was called Advanced Proteomic Platforms and Computational Sciences. In this program, researchers were encouraged to create new experimental techniques or optimize existing ones. The resulting platforms will be high-throughput and ultimately should perform better than current techniques. New hardware and software to analyze clinical cancer samples also were covered by the RFA. “We’ve asked the community to develop new sorts of programs, applications, and capabilities that could provide better accuracy and better precision” for MS, explains Henry Rodriguez of NCI, who is a program director for CPTI. At the same time, NCI asked researchers to develop software that compares MS data sets with other complementary platforms, such as affinity-based technologies.
Comparing existing technologies for reproducibility was at the heart of the second CPTI RFA. Applicants for the Clinical Proteomic Technology Assessment for Cancer program proposed plans for how they would systematically determine the sources of variability for at least two MS platforms and one affinity-capture platform using up to three sets of cancer samples. Rodriguez says that once the results from this program are analyzed, “You’ll really see an acceleration of proteomics from being just a research-oriented tool or application to one that can be used in the clinic and truly give the kinds of answers that clinicians are looking for.” Because the awards are cooperative agreements, investigators will work closely with each other and with NCI staff to share resources and to ensure that certain milestones are achieved.
To facilitate collaboration and the exchange of data and ideas, awardees from both programs will use NCI’s cancer Biomedical Informatics Grid (caBIG). Anna Barker, deputy director of NCI, describes caBIG as “the World Wide Web for cancer research”. caBIG is an open-source, open-access information network that allows researchers to share large amounts of data and huge software files. Because of the nature of the grid, researchers can use these products remotely without having to download them. All tools and data developed by CPTI awardees are required to be caBIG-compatible.
The third CPTI component is the development of standard proteomics reagents. The Clinical Proteomic Reagents Resource program is not accepting applications yet, but Rodriguez says that these awards will be made as contracts to various groups to develop and characterize reagents, such as proteins, peptides, and antibodies, which will be made available to the entire cancer community.
All of these CPTI programs are significant efforts that will involve many institutions in a multidisciplinary network, says Rodriguez. NCI administrators hope that proteomics researchers, by working together, can really start generating reproducible data that will accelerate biomarker discovery and lead to new tools for patient diagnosis and treatment.
