OpenWetWare:PLoS community page
|Discussions||Science 2.0/Brainstorming||→||→||Science 2.0||→|
|Lab notebooks||IGEM:IMPERIAL/2006||→||→||"Engineering a molecular predation oscillator" in IET Synthetic Biology||→|
|Labs & people|| Kafatos:Research|
Fotis C. Kafatos
|Protocols||DNA ligation||→||DSpace||→|| "Directional Cloning into Plasmid Vectors"|
in Molecular Cloning (subscription only)
|→|| Talk:DNA ligation,|
Protocol Online Bioforum
Yeast Pheromone Model wiki
RCSB Protein Data Bank
|→|| E. coli genotypes,|
|Courses||20.109: Laboratory Fundamentals of Biological Engineering||→||→||"20.109" on MIT OpenCourseWare||→|
|Articles||OpenWetWare:PLoS community page||→|| (insert dspace/nature link to PLoS preprint),|
|→|| (insert link to PLoS Biology article) in PLoS Biology,|
|→|| OpenWetWare_talk:PLoS community page,|
PLoS ONE commentary
Caption: OpenWetWare fills current communication gaps in the scientific pipeline. Existing communication outlets focus on publication whereas OpenWetWare fills both the pre-publication and post-publication needs of scientists. Specific example pages from OpenWetWare are given (normal font, pages can be found at http://openwetware.org/wiki/XXX). For a complete listing of links, please see XXX page on OpenWetWare.
OpenWetWare (http://openwetware.org) is an effort to facilitate the open communication of scientific information including methods, results, educational materials, discussions and publications. In research, the traditional mechanisms for sharing work are reference books, journal papers, personal communications via conferences and invited talks. OpenWetWare seeks to address limitations in these communication methods, by decentralizing and lowering the barriers to information exchange among all researchers, be they professors, students or research scientists. In addition, the goal is to foster a culture in which researchers openly share their experiences thereby reducing needless duplication of effort and increasing the pace of scientific discovery and technology development.
Problems with the current publication system
There currently exist huge inefficiencies in the process of dissemination of research. Here we outline a few of the problems with the current publication system that OpenWetWare seeks to address.
Journal articles fail to capture much of the information generated during the research process.
Journal articles are the most common way that researchers disseminate their work. Yet these articles are only written once a project is near completion and are, by necessity, limited in length. As a result, authors omit much of the information generated during research, such as alternative hypotheses, failed experiments, protocol details, and raw data. Either these details are forgotten during the writing process or there is no space for them in the final article. Hence, much of the knowledge from a research project never finds its way into a formal publication medium and is easily lost forever. Instead, this work remains buried in lab notebooks or is passed on as collective wisdom between successive generations of students in a lab. These gaps in published information creates barriers for other researchers to replicate a result. Even within a single lab, there is the challenge of maintaining knowledge as people leave and enter the group.
Information sharing in research is slow and limited.
The lag time from final experimental result to appearance online of the finished publication can easily be months or years. Personal communications are faster at spreading results than publications. However, they are inefficient at disseminating information widely and fail to take advantage of the democratic and decentralized forms of communication made possible by the Internet. For many students and research scientists below the principal investigator level, there are scarce opportunities to discuss their work with a broad audience. Because of time constraints, a principal investigator necessarily cannot discuss the results of every lab member during talks and personal interactions. As a result, information is shared inefficiently and many potential opportunities for collaboration are lost.
OpenWetWare provides efficient means for researchers to disseminate information about every piece of data generated during the research process in more detail and much quicker than traditional communication methods. We seek to embed open sharing in the research process by making it easier for researchers to do the things they already do. OpenWetWare uses a wiki as a platform for capturing and sharing biological knowledge as it is generated. As seen by the success of Wikipedia, wikis offer a simple, collaborative, and online editing environment. We are always seeking new tools and technologies to make open sharing of research easier. To further promote sharing and reuse of the information on OpenWetWare, all content is dual-licensed under the Creative Commons Attribution-ShareAlike 3.0 Unported license and the GNU Free Documentation License version 1.2. Daily dumps of the entire site are also available for download.
Experiences with OpenWetWare to date
Our experience with OpenWetWare has been overwhelmingly positive. The flexible nature of a wiki allows individuals and labs to use it however works best for them. A sampling of different ways in which researchers are making use of OpenWetWare is listed in Table 1 and discussed below.
A lab's webpage ideally communicates the ongoing research of the lab to the outside world. In reality, maintaining a lab's webpage generally falls to one person and the content quickly goes stale. OpenWetWare provides labs with an easy way of generating a "normal" non-wiki lab webpage from the wiki. This permits all lab members to easily edit the lab's webpage, improving the chances that the webpage reflects the current lab research.
The experimental protocol is one of the most critical types of information. Yet with journal space limitations, materials and methods sections are often the first to get shortened. This trend makes it more difficult for others to reproduce published results. Although journals dedicated to methods exist, most researchers don't publish their protocols either because they believe the protocol development does not warrant a separate publication or because the effort needed for publication is too great. In addition, revisions or failed protocols are essentially never published. Often, knowing the process by which a researcher decided on a particular protocol is enlightening. By providing a low cost for posting protocols, OpenWetWare is again capturing knowledge that would otherwise never be shared or digitized. The wiki also provides for real-time discussions about protocols. For example, when Sean Moore, a member of Bob Sauer's lab at MIT posted his protocol for β-galactosidase assays, his labmate Kathleen McGinness then annotated the protocol with sample results she obtained while using it . Thus, archiving protocols on OpenWetWare allows for feedback by others. Over time, this process should give rise to a set of reliable and widely-used protocols. We are also working on developing curated consensus protocols .
The lab notebook is the most detailed record of what a researcher does and what they learn in their work. Therefore, any effort to capture all knowledge from the research process must foster sharing of the lab notebook. Rather than allowing information to languish in paper lab notebooks, OpenWetWare encourages users to document their work digitally online so that the work is searchable and browsable. OpenWetWare-based lab notebooks have gained the most traction with the teams of undergraduate students participating in the international Genetically Engineered Machines (iGEM) competition (http://igem2007.com). Many teams in the competition opt to document their entire projects on OpenWetWare. For example, the 2007 Imperial College London iGEM team included their brainstorming session, described their chosen team projects, kept their daily lab notebooks, listed their cloned DNA constructs, summarized modeling results and more . The easy-to-use and collaborative nature of the OpenWetWare wiki allows all the students and their advisors to stay apprised of how the work is progressing. In short, the Imperial College London iGEM team's notebook is a complete, digital, online record of their research project complete with timestamps and author attribution. It is the model for how embedding of sharing and digitization of biological knowledge in the research process should work.
MIT has published virtually all of its course materials online via OpenCourseWare. Inspired by OpenCourseWare's example, we have also encouraged professors to host their courses on OpenWetWare. OpenCourseWare is similar to a journal article in representing a final published version whereas OpenWetWare provides a place for pre-publication and constantly updated content. Importantly, the collaborative editing interface of the wiki provides a forum for students, teaching assistants, and professors to interact and discuss course content.
For example, MIT's Department of Biological Engineering runs its introductory lab techniques course on OpenWetWare . All background material, protocols, and homework assignments are posted on OpenWetWare. As students follow the protocols in lab, they can correct any errors or points of confusion in the lab manual. They can also post their results online to share with their classmates. Running the course on OpenWetWare engages the students more directly in the course. Rather than being passive consumers of the course materials, students are active contributors. Revision histories ensure that instructors can easily monitor changes to course pages.
Publishing all course information not only benefits the students in the course. Detailed class protocols intended for novice experimentalists are especially useful for newcomers to biological research. Also, educators at other institutions can easily draw upon the course content in their own classes (while giving proper attribution as required by the copyright license).
Initiatives in development
In keeping with our goals to lower the technical barriers to sharing research, we have also begun a number efforts in scientific publishing. In particular, we have launched an initiative to compose review articles on OpenWetWare (http://reviews.openwetware.org). Wiki-based reviews should enable communities to collaboratively summarize and document the latest work in their field . We are also piloting the introduction of OpenWetWare blogs for those who prefer a more diary-like interface for expressing their work .
Challenges for the future
We have encountered researchers with reservations regarding the site. Here we outline some of the most common concerns.
- Some are concerned about vandalism of their postings. However, when they learn that every user account is attributable to an individual with their full publication name, these concerns are allayed. Moreover, OpenWetWare offers several mechanisms to easily track changes made to pages of interest.
- Some researchers value the barriers of the traditional publication process as a convenient means to limit the flow of extraneous and potentially unreliable information. However, we feel that search technology will improve as the digitally available commons of biological knowledge grows. The first step is to make sure scientific information is digitized rather than be lost forever. Just as researchers develop new tools to mine information from the wealth of genomic data, researchers are developing tools to mine information from the web.
- Some scientists have a resistance to posting research progress and topics prior to their publication due to fear of being 'scooped'. This concern derives from academic community norms in which researchers only receive credit upon publication of the final research article. We hope to start to change these norms by demonstrating that sharing information about research in progress is far more likely to lead to collaboration than to competition.
Although OpenWetWare has been very successful to date, there is still much to be done to advance the open sharing of research in biological science and engineering. In particular, until the open sharing of research is recognized in the traditional reward structure in science, initiatives like OpenWetWare will remain on the fringes of most fields. We need to construct a publishing pipeline in which sharing and digitization is integral to the process of research: from project conception to final publication. Researchers need to "receive credit" when they make their protocols, datasets, model files etc. freely available to others. We consider this to be a critical but long-term goal of OpenWetWare. In principles, many of the key mechanisms are in place for this to happen on OpenWetWare. First, all new users on OpenWetWare are assigned a username based on their full, publication name. Such a convention more closely aligns OpenWetWare with journals and conferences in which use of your full, real name is the norm. Second, every update to the wiki has a timestamp and is attributable to an individual under their full publication name. Thus, in the future, it should be possible to give credit to researchers for contributions beyond papers and talks. What we lack is a means to go from documented research project on the site to a peer-reviewed journal publication. Constructing this complete pipeline is a key step in promoting the open sharing of research.
In just over two years since OpenWetWare's inception and relying primarily on word of mouth advertising, the site has over 3200 registered contributors and more than 100 academic labs from around the world. We invite all researchers in biology and biological engineering to join the OpenWetWare community. To begin contributing, please visit http://openwetware.org/wiki/Join and request an account.
OpenWetWare's past and continued success is entirely due to the efforts and passion of the entire OpenWetWare community. We thank them for their invaluable ideas, contributions and support.
Funding. The authors acknowledge funding from the U.S. National Science Foundation Biological Databases and Informatics program award 0640709 to Drew Endy, Department of Biological Engineering, MIT.