“Team Science” is necessary when a project is bigger than one narrow specialty can accomplish on its own. Over the last few years, I was involved with a project from National Cancer Institute exploring the proteogenomic differences among colon cancer tissues. The project was part of a U24 cooperative agreement, which implies expected cooperation among multiple institutions, not just laboratories. Our work, which was recently published in a major journal, illustrates why Team Science is an important paradigm for researchers.
Major biomedical advances depend upon collaboration. The days when a major research paper was envisioned and executed by an individual researcher are gone. An individual laboratory may make an advance marking a collaboration between an investigator and his post-docs, graduate students or staff researchers. Generally, though, the papers that mark the greatest accomplishments span many such laboratories. The reason is that researchers have been compelled to specialize along the way; the number of people who can claim real excellence in multiple disparate fields is quite small.
In the case of our colon cancer work, our cooperative agreement was leveraging work accomplished by another major effort called The Cancer Genome Atlas (TCGA). TCGA had already acquired these colon cancer biopsies from medical repositories. They subjected the samples to extensive genomic, epigenetic, and transcriptomic analysis, producing significant papers on these features of the tumors. My cooperative agreement, the Clinical Proteomic Tumor Analysis Consortium (CPTAC), was to add proteomic characterization for the samples and connect those data to the genomic characterization from TCGA.
Within CPTAC, these goals spanned a host of specialties. For example, the network contracted with Enterprise Science and Computing to establish a Data Management and Coordinating Center for the network, enabling us to pass files between institutions and to produce public repositories of data. We have an extensive effort in build sample repositories for later researchers. The teams from individual institutions span biochemists (for sample preparation and enrichment), analytical chemists (to plan and execute mass spectrometry experiments), bioinformaticists (to translate the data to information), and biostatisticians (to evaluate variability, determine reliable differences, etc). This places a substantial premium on the ability of these scientists to collaborate effectively.
One of the cornerstones of collaboration is respect. I know that scientists do not have the best reputation for social graces, but in fact we have to practice these skills every day. I once heard a genome researcher scoff that “proteins are just DNA’s way of making more copies of itself.” Obviously this comment would not encourage a proteome researcher to seek him out as a research partner. Understanding that the statistician has something key to offer your work is the first step to recruiting him or her.
One of the results of collaboration is broader knowledge. If I have been working closely with an analytical chemist, I will probably understand more about analytical chemistry from that partnership. This has been one of my key advantages as a scientist. Because I received my Ph.D. training in bioinformatics while I was part of a mass spectrometry laboratory, I understood the chemistry much better. As a result, I was better able to understand what analytical chemists needed and communicate that to people with computational training. The flow worked in the other direction, as well; when a computer scientist had a particular skill to offer, I was better able to figure out how it could be placed in the service of analytical chemistry.
I know that there are many scientists who do not particularly enjoy the meetings and teleconferences that seem to dominate our schedules, but these communications generally represent potential points of connection for us as scientists. That network of lab-to-lab relationships is what will move biomedical research forward.