While data science is already contributing to scientific discovery, substantial systemic challenges need to be overcome to maximize its impact on academic research. This is one of three grants, made as part of a five-year, $37.8 million partnership with the Gordon and Betty Moore Foundation, that aim to empower natural and social scientists by strengthening the ability of select U.S. colleges and universities to successfully conduct data-rich and computationally intensive research. Over the next three years, supported campuses will use grant funds to develop meaningful and sustained interactions between disciplinary researchers in the natural and social sciences (e.g. astrophysics, genetics, economics) and researchers in the methodological fields that deal with large scale data collection and analysis (e.g. applied mathematics, statistics, computer science). In addition, supported campuses will establish long term, sustainable career paths for data scientists, and develop an ecosystem of analytical tools and research practices that will facilitate effective research across a range of diverse scientific disciplines. Additional funded activities include holding workshops and training sessions for scientists who work with data, identifying data-science bottlenecks faced by researchers, and disseminating lessons-learned to the academic and research communities.

While data science is already contributing to scientific discovery, substantial systemic challenges need to be overcome to maximize its impact on academic research. This is one of three grants, made as part of a five-year, $37.8 million partnership with the Gordon and Betty Moore Foundation, that aim to empower natural and social scientists by strengthening the ability of select U.S. colleges and universities to successfully conduct data-rich and computationally intensive research. Over the next three years, supported campuses will use grant funds to develop meaningful and sustained interactions between disciplinary researchers in the natural and social sciences (e.g. astrophysics, genetics, economics) and researchers in the methodological fields that deal with large scale data collection and analysis (e.g. applied mathematics, statistics, computer science). In addition, supported campuses will establish long term, sustainable career paths for data scientists, and develop an ecosystem of analytical tools and research practices that will facilitate effective research across a range of diverse scientific disciplines. Additional funded activities include holding workshops and training sessions for scientists who work with data, identifying data-science bottlenecks faced by researchers, and disseminating lessons-learned to the academic and research communities.

While data science is already contributing to scientific discovery, substantial systemic challenges need to be overcome to maximize its impact on academic research. This is one of three grants, made as part of a five-year, $37.8 million partnership with the Gordon and Betty Moore Foundation, that aim to empower natural and social scientists by strengthening the ability of select U.S. colleges and universities to successfully conduct data-rich and computationally intensive research. Over the next three years, supported campuses will use grant funds to develop meaningful and sustained interactions between disciplinary researchers in the natural and social sciences (e.g. astrophysics, genetics, economics) and researchers in the methodological fields that deal with large scale data collection and analysis (e.g. applied mathematics, statistics, computer science). In addition, supported campuses will establish long term, sustainable career paths for data scientists, and develop an ecosystem of analytical tools and research practices that will facilitate effective research across a range of diverse scientific disciplines. Additional funded activities include holding workshops and training sessions for scientists who work with data, identifying data-science bottlenecks faced by researchers, and disseminating lessons-learned to the academic and research communities.

While citizen science projects, crowdsourcing, and other forms of mass collaboration on the Web hold the promise to contribute significantly to scientific research, they often lack adequate institutional or systemic controls to properly mitigate data privacy, cybersecurity, legal, and financial risks. Without such controls in place, government entities or other large institutions are often barred from collaborating with citizen science initiatives, limiting their usefulness and impact. This grant supports efforts by the Commons Lab at the Woodrow Wilson International Center for Scholars to help reduce these barriers by identifying, assessing and prioritizing the risks associated with mass collaboration projects and developing standards, policies, best practices, and other resources that both government agencies and citizen entrepreneurs can use to work together more effectively. Over the next two years, the Wilson Center will publish two peer-reviewed journal articles on privacy, human subjects, and intellectual property issues; host a roundtable series on cybersecurity; construct an inventory of U.S. government involvement in mass collaboration projects; hold a policy briefing for government agencies; and analyze governance models for mass collaboration projects.

In many scientific fields, the process of cleaning and preparing data is managed by increasingly well-established software pipelines. Raw data goes in and structured, refined data comes out ready for analysis. Data pipelines are particularly valuable in fields where the data coming out of instruments are relatively standardized—genomic sequencers, for example, or astronomical telescopes. One potential benefit of data pipelines is that they lower the barriers to sophisticated data visualization, since platforms to explore data visually could be directly connected to data pipelines rather than rely on costly work by individual researchers to prepare and load data. Yet while basic visualization capabilities have been hard-wired to specific data pipelines, there is no generic framework that could interface between data pipelines and data visualization tools.This grant supports efforts by biologist Holly Bik of the University of California, Davis to develop just such a framework. Partnering with leading data visualization firm Pitch Interactive, Bik will work with an initial set of use cases to develop a framework for data visualization on top of existing genomic data pipelines, keeping an eye toward its applicability to other fields.

Three open source programming languages form a canon of sorts for the emerging field of data science: Python for general computation; Hadoop for managing massive unstructured data; and R for statistical analysis. Funds from this grant support efforts by Karthik Ram, a postdoctoral scholar at the University of California, Berkeley, to expand and strengthen the R community through the development of products aimed at lowering the barriers to the use of R. Ram has developed an R software module, for instance, that greatly simplifies the process of gathering data from archives and services commonly accessed by scientists, like Dryad, the Global Biodiversity Information Facility, or the Biodiversity Heritage Library. Ram’s module thus obviates the need for scientists to write their own idiosyncratic code to parse data from such repositories. Grant funds will support the further development of R modules by Ram and his team, as well as outreach efforts to the scientific community to provide training and speed adoption of the new tools.

This project funds an ambitious project by Brian Nosek, a professor of psychology at the University of Virginia, to develop and expand an institutional framework for collaborative scientific work—the Open Science Framework (OSF)—that’s modeled on the open source development protocols that have been so successful in the cooperative development of software. Nosek’s project is based on the insight that scientists could develop much more efficient collaboration practices, saving themselves time and improving the quality and velocity of their work, by borrowing the basic methods and tools of open software development. These include versioning (creating an edit log that tracks changes to any files associated with a project), “tagged releases” (locking a particular, tested version of a project for broader dissemination), “forking” (creating a personal copy of a project to add one’s own edits or additions), and “pull requests” (a request to the owner of a project to merge changes in a “forked” version back into the original). Funded activities include further development of the OSF, the construction of an applications programming interface that would allow the OSF to seamlessly interoperate with other tools and platforms, and collaborations with other developers of scientific cyberinfrastructure.