
Six years ago, Harvard scientist Jay Bradner discovered something unusual. His laboratory had isolated a molecule in mice, named JQ1, that appeared to reverse the effects of a serious cancer. But what he did with JQ1 was even more unusual: Rather than submit the findings to a prestigious journal, Bradner openly distributed the structure of the molecule — free and reusable for anyone.
Bradner’s goal was to disseminate his lab’s finding as widely as possible to encourage and expedite collaborations. And although the original concept for JQ1 fell through, he and others ultimately succeeded on different fronts: JQ1 now has broad usage in treatments for HIV infection, heart disease, pancreatic cancer, and more.
Despite this success story, most scientific research today is not published openly — meaning freely, to everyone, without delay from the time of publication. Instead, it lives behind time embargoes and paywalls, costing as much as $35 per article to access. Even when scientific information is free to read, it is subject to copyright restrictions that prevent it from being recast quickly in new ways.
A growing movement for open access seeks to change this because limitations on the use of scientific discoveries hinder the efficiency of research, increase costs, and ultimately delay or even impede scientific progress. The possibilities of open access could be transformative.
If published research and data were freely accessible and reusable by researchers of diverse interests, urgently needed solutions could be greatly accelerated. Scientists could quickly cross-check important studies, catching potentially consequential mistakes. Medical providers could access the latest technical guidance, improving patient care. And students around the world could build on each other’s work. With openness, good ideas could truly come from anyone, anywhere.
Some early open access efforts — like the Human Genome Project, one of the world’s most ambitious programs undertaken to sequence the complete set of DNA in a human body — illustrate just how significant the impact of open access could be. The Human Genome Project shared each of the body’s three billion DNA letters freely and rapidly online, leading to the discovery of more than 1,800 disease genes and generating over $1 trillion in economic impact across a diversity of sectors, including health care, energy and agriculture. Critical to this was enabling all uses of the data by anyone for any purpose — academic or commercial.
This is the future envisioned by a growing number of institutions and universities that sponsor research and development. In the last few years, NASA, Research Councils UK, and Wellcome, as well as the Massachusetts Institute of Technology and the University of California system have all implemented open access policies.
The Bill & Melinda Gates Foundation will usher in 2017 by making all published research funded by its grants available on full open access terms. It will base its open access policy on a robust legal framework that supports full reuse rights — the Creative Commons Attribution-Only license, also known as CC BY. This license, which is also applied to all materials authored by librarians and staff of the University of Michigan Library, lets anyone distribute and build upon a work’s underlying data as long as they credit the original creation.
CC BY facilitates processes such as text-mining, helping researchers understand patterns in large datasets. It also promotes innovation, unlocking the ability of people across sectors and geographies to build on one another’s work. Unlike its counterpart, the CC BY-NC license, CC BY permits commercial reuse, meaning it directly enables the kinds of public-private partnerships that are so essential to scientific innovation.
That’s why, in part, the Gates Foundation adopted it — because it permits the widest range of actors to do their best work on the widest set of problems.
To be clear: free and open reuse does not mean misuse. Under CC BY, researchers and publishers do not give up the value of their name: The license includes significant mechanisms to guard against academic misrepresentation, and protects the logos that journals use to brand their reprints and products.
CC BY is a growing license of choice among open publishers, including the Lancet and Cell Reports. It is the top choice for researchers. In 2015, the Nature Publishing Group reported that 96 percent of researchers publishing open access selected the CC BY license.
Changing the status quo is tough, but it is made possible when people come together to reconsider outdated assumptions and practices. With the support of researchers, publishers, and the academic community more broadly, legal frameworks like CC BY can expand the interconnected web of human knowledge and facilitate its use by everyone.
In a fast-changing world, making these adjustments will advance what’s best about the scientific process: the ability of humans to generate new lifesaving and life-changing ideas by building on the work of the past in a world where sharing and using data in medicine, public health, and the environment is the norm rather than the exception.
Richard Wilder is the associate general counsel at the Bill & Melinda Gates Foundation. Melissa Levine is the lead copyright officer at University of Michigan Library and ex officio member of the library’s open access committee.
I agree that Open Access publishing would be good for expanding access to biomedical information.
The article could do a better job differentiating between open access (openly licensed journal articles) and open data (openly licensed data). The examples are mostly focused on the benefits of open data, but much of the narrative is about the importance of open access publishing. It’s dangerous to conflate the issues, since scientists become confused and assume that their publication of a study automatically makes the underlying data “open” (https://doi.org/bndt).
Note that the Gates Policy specifies publications must be CC BY 4.0 and data must be “Accessible and Open Immediately.” I applaud the Foundation for not overlooking the importance of open data. It would be great to specify a default data license of CC0 (public domain dedication). Also it would be great to specify that code must be released under a permissive open source license.
Cheers to Gates for leading the way on these pressing issues.