This document is a work in progress to describe the advantages of using a wiki as an information repository and collaborative space for institute research labs.
What good are wikis in general?
Wikis make it easy for anyone in a group to contribute without an administrative bottleneck. There is no "webmaster" who approves every contribution to the site. Thus, content is created and edited much faster than is possible otherwise. There are now ~1700 pages in OpenWetWare over the past few months. This is quite remarkable given the small community that is contributing.
Ease of contribution
All you need is a web browser (and a login/password in the case of OpenWetWare) to contribute to a wiki. The editing language is meant to be extremely simple and easy to learn.
Hard to break
Wikis are nearly impossible to break - changes are tracked and can be very easily reverted. The entire history of a page can be seen quite easily.
Why are wikis useful for research labs?
The amount of information and expertise that is accrued in labs is tremendous. Conveying this information and expertise to subsequent lab members and to the greater scientific community now often relies on talking and interacting closely with the right people. One reason for this is that it is quite difficult to put these thoughts down in a systematic way. Tremendous amounts of individual expertise and information is lost when lab members leave (a key problem given the high turnover rate of academic labs). The wiki provides a low-barrier of entry method for lab members to contribute that information to a database which will persist after they leave the lab.
Often a method, such as a lab database, is created that is not evolvable enough and has too many bottlenecks to be useful to lab members. At this point, the system is usually abandoned, or becomes very minimal (i.e., this is where we keep our freezer stocks.) The dynamic nature of wiki pages allows the structure to form into the most useful way to convey information. The ease of wiki linking enables richer information sources than is possible with static documents. For instance, a user can link out to informative pages about particular words in a wiki page, enabling someone reading the document who doesn't understand a concept to quickly locate a reliable, accurate definition. Moreover, since everyone can revise the information content of a wiki, mistakes are more quickly caught and corrected.
By providing a common space for people to post information about their work, graduate students are more likely to be aware of the work going on in other labs locally. It seems like this will improve the likelihood of collaboration, and also a provide a source for finding out where certain expertise lies.
Shared Materials Information
Chemicals, vectors, strains, antibiotics, computing, etc., can have their own wiki pages listing general information, safety, and more. Since these are things which are general information rather than lab-specific (like protocols) they could be shared across labs. For example if a protocol called for the use of Ampicillin, it could include a link to this, so if you forget the mode of action of ampicillin you just click rather than have to go dig it up in Molecular Cloning.
A list of equipment provides a useful shared resource to make labs aware of the available local equipment. Also, equipment pages serve as a central repository for control experiments and other information about the device. For example, the Victor3 plate reader located in the Endy Lab has posted not only simple usage information, but also sets of controls on a variety of topics such as how the lamp energy affects signal to noise ratios.
Why not just start your own wiki?
Assuming that you agree that wiki's are a good way of sharing information between lab members, the next natural question is why join OpenWetWare? Why not just start your own lab wiki?
Reasons to join
First, wikis work when the number of excited initial users create the beginnings of an useful resource. Other users begin realizing how useful the resource is and begin contributing. We feel that OpenWetWare has already begun this. Joining OpenWetWare will probably get more people excited than otherwise would be, therefore increasing the probability that the lab will make useful contributions.
Second, the shared resources sections have the potential to be much better than with any lab alone. Since most labs have many of these shared resources in common, it behooves us to collaborate on their content generation and upkeep, rather than repeating that work over and over again.
Third, wiki's also require some backend users that upkeep not only the content, but the look and feel, and overall organization. Currently we have a few people that do this in the backgroud, that have led to current state of OpenWetWare. However, if you ask any of them, there are not nearly enough people for all the work that could be done. Starting a lab wiki from scratch would probably require some initial effort on the behalf of these people. However, tapping into a community that already exist would probably reduce the amount of effort compared to starting from scratch. Also, it would provide us with more of these backend users :-).
Fourth, there are collaborative sections that will be better the more labs that join. For example, the Protocols section provides a general area to post protocols. While a protocol describing how to perform in vitro DNA ligation is useful from one lab, it is probably even more useful for many. The DNA ligation protocol just describes a general method, while the Endy Lab and the Knight Lab have their own protocols and reasons to do them differently. A balance between shared and lab-specific protocol space allows for optimization of protocols and faster learning curves for new lab members trying to learn techniques.
These concerns are quite common, and important enough to list out here.
Our own space
Many labs feel as though their presence will be diluted in a larger site such as OpenWetWare. This is a concern we share to some extent. For example, both http://web.mit.edu/endy and http://www.syntheticbiology.org both point to pages within OpenWetWare, and this arrangement may have less branding effect than if the sites were independent. We have found the collaborative and dynamic nature of OpenWetWare to be worth the loss in brand. In addition, what we have been finding is that there are ways to make the navigational experience for the user more branded. For example, most Endy lab pages have a navigation bar template that allows visitors to traverse the Endy lab pages easily, without being too distracted by all the other content available. We are currently working on ways to customize this further (for example, the ability to change the sidebar). If you have any ideas, let us know.
Another reason people want their own space is that someone from the Endy Lab doesn't necessarily want a Grossman lab ruffian to edit their protocols. To combat this to some extent, we have been labeling lab-specific pages (pages that a lab doesn't want edited by other labs) by prefacing them with the lab/organization name (for example Endy:Research). As a matter of etiquette, everyone is asked to be respectful of pages that are not part of the user's organization. We have not had problems with this so far.
One of the biggest concerns with wikis is the ability for anyone to vandalize other people's work. While sites like wikipedia are able to deal with it because they have such a large benevolent user base compared to the number of vandalisms. The community we are creating with OpenWetWare is significantly smaller. Thus, we can't assume that vandals will be caught quickly. Therefore, we have decided to make it such that only registered MIT users and groups can edit content. That way, each edit is trackable to a specific person, which can be tracked down by their email address. We find that this public embarrassment factor, while increasing barriers to entry, provides sufficient impetus to prevent malicious or inconsiderate editing. So far, we have had no problems with vandalism, or unwanted edits, whether malicious or otherwise. Thus, we feel that this system is working. We will re-evaluate if and when a single instance of a problem arises.
One concern that people often have is if they are part of OpenWetWare, all their information will be publicly available. We realize there is an opportunity cost in not allowing posts of private information, which could be more useful to certain labs, but won't be posted for secrecy/competition concerns. We have considered making a more private site and a corollary open one. Right now this is very difficult. In the future, once permissions are added to the MediaWiki software, we will reconsider making private areas. Currently, some people in the Endy Lab run wiki's on their private computers to store their lab notebook and individual project data. For the vast majority of postings so far, there is no reason to make them private.
Generally, we feel the tendency is for labs, particularly in the biological sciences, not to share information in a generally available way. (Most labs are very helpful on an individual, case by case basis however.) But if everyone shares their expertise then ultimately everyone benefits from the "open commons" of information. There is also often a general concern that by making ideas/expertise publicly available, such as on a website, you risk someone else either a) scooping your work or b) somehow losing your edge against other labs. Usually, however, if you are open and sharing of information, then others will respond in kind with the long-term benefits outweighing short-term disadvantages. Two proofs of this principle are the open source software community and wikipedia. Alternatively, there is the purely pedagogical reason that a fundamental goal of all universities is education: we have an obligation to share our information in any way possible.
How do I join or I have a question?
Email endipedia-admin AT mit DOT edu. Feel free to email regardless of whether you are an individual, a student group, a lab or <insert your group here>. Your research or interests do not necessarily have to be biologically relevant either.
The current usage of OpenWetWare provides some compelling examples of why joining the site may be useful to your lab.
Wikipedia is a great example of a useful information resource being created via a wiki.
See the Interwiki map for other wiki's.