Data sharing in response to emergent health crises: an interview
with Dr. James Crowe
Whether describing an emergent disease, developing a vaccine, or selecting a targeted
therapy, intra- and interinstitutional collaboration is vital to expeditiously tackling urgent
biomedical questions. However, many characteristics of academic science restrict scientists’
ability to work cohesively. First, distributing limited financial and biological resources nationally
or globally presents a substantial logistical challenge. Additionally, labs and institutions compete
for funding within a scientific niche, which requires them to demonstrate productivity in the form
of peer-reviewed publications. Peer review, the cornerstone of science, can take months from
paper submission to an accessible publication, many of which are only available through
personal or institutional journal subscriptions. What strategies is the scientific community
implementing to overcome these barriers to improving human health?
The Zika virus response has implemented various policies to facilitate rapid data
sharing. Numerous journals are pledging to provide open access to Zika-related manuscripts
and datasets. In February of this year, the World Health Organization (WHO) released
guidelines to readily consolidate emerging data about the Zika virus into a centrally located and
freely available hub. Scientists are also establishing networks to share genomes and reagents
so that instead of building tools, labs can expend energy investigating critical questions, such as
pathogenesis of Zika infection. At Vanderbilt University, Dr. James Crowe and his lab used Zika
virus sequence information that became available quickly to synthesize genes that encode the
viral protein to be used for antibody discovery.
This is one example of how open data is benefiting research on Zika virus and its
associated diseases, leading public health officials to praise scientists for fulfilling their
“fundamental moral obligation” to collaborate. However, this movement has also been criticized;
authors of a Nature editorial warned that depositors of raw data will not receive adequate credit.
Undoubtedly, the lack of established norms for sharing and using unpublished data and the
departure from traditional peer review pose challenges to open data sharing. We asked Dr.
Crowe for his perspective on these obstacles and how to overcome them.
What are some challenges that have arisen or could potentially arise from open data?
Historically, publishing in high quality publications is the key to academic success and indirectly
to grant funding. Public release has typically killed the ability to publish high impact findings in
high visibility journals in the past because of press embargoes. The situation is evolving, but
promotion and tenure and grant review processes haven't changed all that much.
What advice do you have for scientists who might want to use data released prior to peer
Peer review really does refine the interpretation of data. I'd be relatively cautious of data that
had never been reviewed.
What advice do you have for scientists who are considering sharing their data prior to peer
If the findings are potentially really important and one felt publication in a high visibility journal
was appropriate, I would interact with the journal staff or editorial board to understand how such
disclosure relates to that journal's policy on prior publication.
The physical sciences have been sharing data openly for many years. Do you see a role for
open data sharing in the biomedical sciences beyond response to emerging crises?
I have never really understood how those fields handle academic credit for promotion and
tenure and for grant funding review. This might be appropriate, but the culture change needed in
the biological fields is enormous and I believe this would take years to accomplish.