“The concept of Open Science stands at odds with the prevailing competitive culture in research today. The goals of Open Science are to create transparency in experimental methodology and data collection; to make scientific data publicly available and to allow this data to be reused by other scientists; and to use publicly accessible and transparent means of scientific communication. In practice, Open Science relies heavily on web-based tools to foster scientific collaboration.”
Open Science: Supporting scientific inquiry, or giving away your work for free?
By Andrea K. Globa
Cell and Developmental Biology, University of British Columbia, Vancouver, BC, Canada
The pressure to “publish or perish” has always characterized academic life, however, this expectation has only intensified in recent years. Academics are under extreme pressure to consistently publish papers in high impact, peer-reviewed journals in order to further their careers (1). Some claim this pressure is beneficial, as it motivates researchers to produce high-quality work early in their careers. However, this increased pressure can also push scientists to self-plagiarism, or to dividing data into a number of short, discrete papers rather than one large, cohesive paper, in order to add a greater number of publications to one’s record (1). Furthermore, the “publish or perish” mentality takes the focus off of teaching and mentorship, both central to the development of the next generation of researchers (1). In order to publish in a high-impact journal, reviewers require extensive data that would be nearly impossible for an individual to collect on his or her own. As a result, the number of authors on the average scientific publication has grown immensely in the past twenty-five years (2). Indeed, in the present day there are few instances of publications with a single author. In particle physics, there are numerous examples of papers with up to 100 authors. Multi-centre clinical trials have resulted in health research publications with astoundingly large author lists (3). In neuroscience, authors are expected to provide data from the level of gene expression, to protein activity at the synapse, to electrophysiology and behavioural measures in experimental animals. In order to be published, a single paper requires the expertise of many individuals.
The concept of Open Science stands at odds with the prevailing competitive culture in research today. The goals of Open Science are to create transparency in experimental methodology and data collection; to make scientific data publicly available and to allow this data to be reused by other scientists; and to use publicly accessible and transparent means of scientific communication (5). In practice, Open Science relies heavily on web-based tools to foster scientific collaboration. The Open Science movement gained ground after mathematician and Fields Medalist Tim Gowers asked some important questions on his blog. He wanted to know if collaborative mathematics was possible using the internet to communicate (6). He posted a difficult, unsolved mathematical question, updated his progress as he worked, and invited anyone to post his or her suggestions and ideas. Many people did indeed contribute. In fact, hundreds of comments were posted over 37 days, at which point the group had solved a more difficult generalization of the original problem (7). Information exchanged was maintained in a public document, which allowed contributors to be recognized when the solution was published in a peer-reviewed journal. Open Science provides a forum for quickly exchanging ideas, providing constructive criticisms, addressing problems and soliciting advice.
In fact, there are academic communities where large-scale collaboration and the sharing of data and resources are an integral part of the group’s culture. For instance, guidelines for the sharing of genetic data have been introduced to enforce data sharing in genetic research. From the outset of the Human Genome Project, all sequence data was to be made publicly available on the GenBank database within 24 hours of being generated (8, 9). This top-down enforcement of data sharing has given researchers access to useful genetic tools. In fact, soon after this policy was introduced, over 30 new genes were implicated in disease (8). This policy prevented the companies involved with the sequencing project from patenting human genes, allowing many research groups to study these genes and determine their respective functions more quickly (8). A culture of sharing also exists within the community of researchers who use the fruit fly Drosophila melanogaster (D. melanogaster) or the nematode Caenorhabditis elegans (C. elegans) as model organisms in biological research (10, 11). The genomes of both D. melanogaster and C. elegans have been sequenced, and this information has been useful in manipulating gene expression in order to determine gene function. Drosophila researchers have access to many information sources through online databases including genome sequence data, proteome sequence data, protein interaction data, homology with human genes, and other useful metrics of the D. melanogaster genome (10). Furthermore, Drosophila researchers can access five international genetic stock centres, where flies with different gene mutations are stored. The largest of these collections contains approximately 20,000 fly strains (10). The ability to quickly access these different stains helps to speed Drosophila research. In the case of C. elegans, when a researcher develops a new knockout strain, they can deposit this strain in the Caenorhabiditis Genetics Center (CGC). The CGC collects and distributes C elegans strains, acting as a central store for strains of this experimental animal. In a 2009 survey of C. elegans researchers, over 60% of respondents had deposited knockout strains into the CGC, and 94% of respondents had requested specific strains from the CGC (11). In particular, the C. elegans Gene Knockout Consortium (GKC) has encouraged the sharing of resources in this community. The GKC is a collaboration among the Moerman group at the University of British Columbia in Canada, the Barstead group at the Oklahoma Medical Research Foundation in the United States, and the Mitani group at Tokyo Women’s University in Japan, which aims to create knockout strains for all known genes in the C. elegans genome, and share these strains with the public before publication (12). This research community is extremely tight-knit, as the field’s founding member, Sydney Brenner, fostered a spirit of collaboration and openness among his graduate students. Many key researchers in the C. elegans field were trained in this environment, and have brought these same values to their own laboratories. As a result, the community seems to be more willing to participate in and enforce Open Science initiatives.
The Gowers’ mathematics example clearly demonstrates the potential benefits of Open Science, and the D. melanogaster and C. elegans model organism examples show how the principles of Open Science can be put into practice in biological research. In examining these few examples of successful Open Science, a pattern emerges. It seems that Open Science is most successful when established researchers prioritize the development of an Open Science initiative. These scientists already have established careers and tenured positions, and thus have little to lose in participating in Open Science. Their participation and leadership can create more opportunities and set a good example for the next generation of scientists, who will have greater resources available as a result of the actions of the established researcher.
There are still many issues to be examined if Open Science is to be adopted by many scientific disciplines. One of the most obvious problems is the issue of assigning and receiving credit. In order for academics to succeed, they must demonstrate that they have the capacity to perform relevant research, and some measure should be used to demonstrate their worth. In the “publish or perish” culture, this takes the form of publication in peer-reviewed journals. Most graduate students compete for limited academic positions at the end of their training, and scientists compete for grants to fund their research. For this reason, many academics are protective of their findings, and would rather avoid the risk of being “scooped” than participate in Open Science. How can young researchers be credited for their work, other than through their publication record? There will also be differences in the willingness to participate in Open Science in academia as compared to industry scientists. Is it realistic to expect that researchers who work for large corporations will share their findings in the same way that academic scientists might? Also, in fields such a pharmaceutical sciences, scientific discovery can make a lot of money for the company or institution that patents the findings. Can we expect researchers who intend to patent the drug they design to share their methodology before it is published and patented?
These questions must be addressed in order for transparency and sharing to become an integral part of research. Although academics may become jaded and distrustful later on in their careers, most enter science with the optimistic notion that they can make a meaningful difference, contribute to their research community, and work with others to develop a greater understanding of the world. Open Science promotes a return to this idealistic view, one that could change the culture of research.References:
- Ushma S. Neill, “Publish or Perish, but at what cost?” The Journal of Clinical Investigation 118 (July 2008): 2368.
- Mott Greene, “The demise of the lone author,” Nature 450 (December 2007): 1165.
- Antonio Regalado, “Multiauthor papers on the rise,” Science 268 (April 1995): 25.
- Daniel Watch, Whole Building Design Guide: Trends in Lab Design, http://www.wbdg.org/resources/labtrends.php (June 18, 2010).
- The OpenScience Project, What, exactly, is Open Science? http://www.openscience.org/blog/?p=269 (July 28, 2009).
- Tim Gowers, Gowers’s Weblog: Is massively collaborative mathematics possible? http://gowers.wordpress.com/2009/01/27/is-massively-collaborative-mathematics-possible/ (January 27, 2009).
- Michael Nielsen, TED.com, Open science now! http://www.ted.com/talks/michael_nielsen_open_science_now.html?awesm=on.ted.com_Nielsen&utm_campaign=&utm_medium=on.ted.com-static&utm_source=facebook.com&utm_content=awesm-publisher (November 2011).
- Eliot Marshall, “Bermuda Rules: Community Spirit, With Teeth,” Science 291 (February 2001): 1192.
- L. Roberts, “Genome Research: A tussle over the rules for DNA data sharing,” Science 298 (November 2002): 1312-1313.
- Kathleen A. Matthews, Thomas C. Kaufman, & William M. Gelbart, “Research Resources for Drosophila: The Expanding Universe,” Nature Reviews Genetics 6 (March 2005): 179-193.
- Matthew R Voell, Lily Farris, Edwin Levy, Emily Marden, “A response to Rome: lessons from pre- and post-public data-sharing in the C. elegans research community,” BMC Genomics 11 (2010): 708-712.
- Donald G. Moerman & Robert J. Barstead, “Towards a mutation in every gene in Caenorhabditis elegans,” Briefings in Functional Genomics and Proteomics 7 (April 2008): 195-204.