A Conversation With
A Conversation with Drs. Christopher Kinsinger and Henry Rodriguez about Sharing Proteomics Data
Last September, NCI's Office of Cancer Clinical Proteomics Research convened a workshop of researchers and stakeholders in the field of proteomics to discuss ways to better share data about the structure and functions of proteins—a challenge facing the entire proteomics community. The meeting, held in Sydney, Australia, addressed establishing standards to ensure the quality of the data, particularly those generated by a technique known as mass spectrometry. A meeting report with recommendations for the field was recently published simultaneously in four journals simultaneously.
Why is open access to proteomics data important?
Dr. Kinsinger: We believe that advances in science can be accelerated when the research community has open access to high-quality data. In coordination with the international proteomics community, NCI is leading efforts to implement policies that govern the adoption and use of quality metrics for open-access proteomic data.
We believe this will enhance the management, integration, and analysis of research data. For example, researchers can use open data to conduct new analyses, which may lead to new insights.
What were the goals of the workshop, and who attended?
Dr. Kinsinger: The goals were to define data quality standards for open-access proteomics data and to build on the Amsterdam Principles. At the Amsterdam meeting, the community coalesced around the need for open access to raw proteomic data from published articles. The intent was to emulate the genomics community's development of the Bermuda Principles of data quality and sharing. The meeting in Australia focused on how we can determine what "quality data" are.
The international workshop included extramural [non-NIH] researchers who produce and use proteomics data, managers of database repositories, editors of scientific journals, and funding agency representatives.
In concrete terms, what did the workshop achieve?
Dr. Rodriguez: The workshop produced specific recommendations for a number of areas, including data annotation; quality metrics and data standards; reference materials; education about data policy; infrastructure for depositing data; and incentives for data sharing such as acknowledging the use of data in publications.
There was also broad agreement about the need for a sustainably funded central repository—or a set of linked repositories—that would work closely with journals to ensure long-term access to proteomics data.
What are the next steps?
Dr. Rodriguez: Several journals have expressed an interest in adopting some of the recommendations and applying them to manuscripts they publish. However, the proteomics field currently does not have stable funding and infrastructure to host proteomics data.
This means that before the journals require authors to deposit data and meet certain quality metrics, there needs to be a reliable place that could host the high-quality raw data.
Can you define "raw data"?
Dr. Kinsinger: Raw data have not been modified beyond the initial processing by the instruments used to create them.
Have you learned from efforts to develop data quality standards in other fields, such as genomics?
Dr. Rodriguez: To move science forward, researchers need and want access to raw data. We realize, however, that often it will take senior investigators and policy-makers to implement policies on depositing data.
Why did you publish the meeting report simultaneously in four journals?
Dr. Kinsinger: This was unprecedented, as far as we know. Based on the overwhelmingly positive feedback they received from meeting attendees and researchers who could not attend, the editorial board of each journal felt it was important to show support for the recommendations. The Journal of Proteome Research, Molecular and Cellular Proteomics, Proteomics, and Proteomics Clinical Applications published the report as a service to the research community.
Were there any disagreements at the meeting?
Dr. Kinsinger: There was some controversy about how data quality will be measured and who would enforce these standards. Many investigators felt that the enforcement of quality standards might stifle innovation in a field that is still rapidly developing technologies.
On the other hand, without standards, no one would know which data to trust.
—Interviewed by Edward R. Winstead