The accuracy and speed of cancer diagnoses are being dramatically improved through a National Institutes of Health funded project that is creating grid-enabled tools that analyze tissue microarrays. Researchers from the Cancer Institute of New Jersey (CINJ), IBM, and several universities are working together in this multidisciplinary collaboration. Cancer research, like a lot of big science, benefits tremendously when many levels of expertise are involved. The project, which received $2.5 million, makes use of pattern recognition algorithms, tissue microarrays, and grid-based computing. Software tools can detect the presence of a specific cancer biomarker from the various tissue plugs on each microarray, and computer vision techniques can locate the biomarker in a specific tissue or sub-cellular compartment, such as the cytoplasm or nucleus. The outcome is a better idea of the most appropriate therapy a specific patient should have based on his or her expression signature. In a proof of concept study, researchers found a very strong correlation between the signatures they generated and patient outcomes and profiles when histologic types and diagnosis were considered. One missing piece is a standardization for naming biological terms and how data structures should be represented in XML schema. The immediate plan is to have a prototype system running in 2008 and a product for the clinical and research communities in four years. Future plans include expanding the number of disorders being investigated and building a clinical support system for oncologists.