Electronic laboratory notebooks aren't just for industry anymore. Many ELNs now are priced for academia, "and rapidly it's coming to the point where science cannot be done without it," says Douglas Perry, professor of informatics at Indiana University in Bloomington.
Large companies typically use ELNs to standardize quality control or establish a legal data trail; academic labs use them to gain searchable access to years' worth of data or the ability to share data easily. But ELNs are not only about moving from paper to electronic notebooks, says Ian Murrell of the Klee Group in France, which produces the Kalabie ELN. "They occupy an important position where knowledge is created out of experiment data, and where collaborative work can be organized."
So, do you need an ELN? Ask yourself these four questions.
1. Do you generate high-throughput or automated data?
Ellen Quardokus, a research associate at Indiana University, had used several sequence packages to analyze the genomics and proteomics data generated in her microbial genetics laboratory. Now, she uses Textco's Gene Inspector ELN to integrate those analyses with experimental designs and results in a unified electronic format. "It had all of the things that I typically do to my sequence in one place," Quardokus says. Since sequences are linked to analyses, when she updates a sequence, results are adjusted accordingly. "It makes it nice to not have to go back and manually redo all of those analyses."
Today even modest academic labs can generate massive amounts of information, thanks to sequencers, microarrays, and other genomics tools. "Getting that data into a lab notebook is either painstaking and error-prone, if you have to write the information in manually, or you cut out the pieces of paper and paste them into your notebook," says Keith Taylor of MDL Informatics.
An ELN can help organize, analyze, and archive these data, Perry says, while also keeping them searchable for later use. It also can help researchers analyze data generated from automated systems in consistent ways, for instance, by automating calculations. "It makes life simpler," says Taylor.
2. Do you collaborate with other labs?
In the mid-1990s, says Jim Myers of the University of Illinois, Urbana-Champaign, "we started realizing that you could use [the Web] not only to get information but to put information out." With funding from the Department of Energy, Myers (then of Pacific Northwest National Laboratory) and his colleagues designed a Web-based ELN (http://collaboratory.emsl.pnl.gov) that allows scientists to do science remotely through collaborations over the Internet.
Collaborating scientists can log in to their system remotely, and then download and upload text, images, and many other types of files, Myers says. Once a researcher submits data to the system, these data are sent out to the server and appear on an HTML page. Users can also use applets to show, for example, three-dimensional protein structures, Myers says. The PNNL program is used mainly by educators teaching biology or chemistry laboratory classes, Myers says, as well as by a few research scientists in small, academic labs.
And it's not just about sharing results: "People can actually put in to the ELN and take out of the ELN in a dynamic sort of way," Perry says. The software can be configured with different workgroups and permission setups to accommodate collaborators, says Christopher Geib, whose Air Force lab uses Rescentris' CERF-Notebook, and "the information can be made readily available to any number of scientists and researchers." For instance, says Rescentris bioinformatician Wolfgang Rumpf, collaborators can annotate entries by suggesting a different protocol, and they can then "actually link their new protocol to my existing entry."
3. Do you generate lots of visual data?
Maryann Martone, neuroscience professor-in-residence at the University of California, San Diego, appreciates information management solutions for modern labs. But as a microscopist, she says, "there was a lot of experimental information that needed to be stored with the hi-res 3-D data, and there was no ready way to do that because the data was hidden in people's lab notebooks."
Martone refers specifically to metadata like protocols, microscope settings, and so on, but the problem could just as well apply to gels, microarrays, and other visual data. Labs typically deal with such information by noting references to image files in paper notebooks, but keeping track of the images, along with experimental information and resulting conclusions, can be challenging.
Martone has found the open-source ELN NeuroSys (http://neurosys.cns.montana.edu/) to be useful, as well as Axiope's Catalyzer. The imaging core facility at Children's Hospital, Boston, also uses Catalyzer to track its data. According to former manager Matthew Salanga, being able to bring images straight from the microscope or other data acquisition instruments into your ELN, where it can be annotated and analyzed, helps research progress efficiently, "because you take a level of error out of it. The raw data is going straight into your notebook."
4. Do you have high personnel turnover?
At the Structural Genomics Consortium in Oxford, UK, researchers archive and track their work using the ConturELN from Contur Technologies. Brian Marsden, principal investigator for research informatics at SGC-Oxford, says the ELN ensures that when people leave the lab, it retains access to their accumulated knowledge.
"There are a number of cases where somebody has left, and we had to go back to look at the data generated previous to their leaving," Marsden says. "In normal cases, we'd have to start from scratch because we wouldn't be able to figure out what that person had done. But we can look at the ELN ? the traces, the gels, the data ? and pick up where they left off," he says.
With about 60 researchers in the facility, Marsden says, communication is of paramount importance even among those who still work on-site. First, it allows principal investigators to know what their researchers are doing. But also, "If someone discovers a new way of doing something, it's easy to communicate what that is," he says: "Set up an E-mail and say, look at this ELN record and you can see what I've done."
With an ELN, a researcher can "search and query notebooks for data, for results, for perhaps experimental protocols that a postdoc did five years ago that you want to replicate," Perry says. Most ELNs allow searches similar to desktop search engines, he adds. Instead of requiring memory of hierarchical structure and organized folders, or even of specific key words, ELN search engines allow previous experiments to be searched by almost any piece of information related to the experiment.