Author Archives: Zac Sarich

Digital Science Internship Reflections

Now that the internship program is coming to a close for me, I wanted to reflect on what the experience has been like for the past year.  If I could describe the entire program with one word, it’d be freedom.  Yes, I did have an overall goal, and I didn’t just come in and do whatever I wanted, but within those confines was a broad area of knowledge that could be tapped into.  In my case, I was dealing with a crossroads between emerging technology, and everyday academic research, with a little bit of design work mixed in.  This required investigating not only how research was done, but probably more importantly, how it could be kept track of in the easiest way possible.  So, most of the time I would find myself researching the practice of research itself, while also looking to see if there was currently anything out there that would fit the bill for a suitable ELN to pitch to faculty members.  Along with this, I picked up a little bit of Adobe Illustrator experience to make some figures, and got to explore ways to design a graphic that will get a message across, while also being appealing to the eye.  The process of having all of these different aspects of a project come together can be pursued differently by each individual, and this is where I think the freedom piece comes in.  I was able to find my own way about these things, instead of being told exactly what to do every day, which enabled me to really explore the things that I found interesting, and whenever I needed guidance or suggestions on what to do next, I was able to get that from Jeffrey.  Because the program was run in this way, I feel like I picked up a lot of useful skills in terms of project design and implementation that will be extremely valuable as I enter the workforce.

ELN Presentation

Since the presentations got cancelled a couple of weeks ago, I thought I might share my presentation in a blog post and summarize what I was going to say for each slide.

Slide02

This outlines some of the major goals of my project.  Overall, I wanted to research some of the electronic lab notebooks on the market that had free trials, with the ultimate goal of getting faculty involved in testing the notebooks and giving me their feedback.  “Implement” is in red here because we’ve created the rubric and have chosen the notebooks, however, we are still waiting for faculty to test them out for us.

Slide03

So what is a lab notebook?  A lab notebook is essentially the main documentation for most labs in terms of results, experiments, and analysis.  They are typically bound, and because they contain so much pertinent information, it is hard to be successful in your research if your lab notebook isn’t well kept.  Furthermore, the lab notebook could be used for things such as publishing your results, as well as proof for intellectual property and patents.  Because your results could be communicated through the notebook, there are specific guidelines that need to be followed to ensure all data is valid.  This includes, but is not limited to, having the notebook be bound, having the page number on each page, and signing/dating the bottom of each page.

Slide04 So, now that we have a general idea of what a standard laboratory notebook should contain, we can talk a bit about electronic lab notebooks.  From here on out I will be referring to electronic lab notebooks as “ELNs” and paper lab notebooks as “PLNs”.  An ELN is essentially just a digitized copy of a PLN that is designed to make research tasks easier.  These tasks range from automated data collection from devices, easy collaboration and communication, and the ability to analyze data quickly and efficiently.

Slide05

Why would we want to use an ELN?  As previously mentioned, it can make it much easier to not only add data to a project, but also track it as the project moves forward.  Because the notebook is digitized, all data that has been submitted to it can be searched and would allow you to search for the name of a reagent you used in an experiment 10 years ago.  ELNs also offer cloud storage that helps to eliminate the clutter that may be caused by a collection of PLNs.  The future of ELNs also looks promising for performing a lot of tedious things automatically such as simple calculations, ordering commonly used equipment/reagents automatically for you, and inserting data such as pictures.

Slide07

Before testing out some of the ELNs, I wanted to visually determine what specific differences were present between ELNs and PLNs.  To do this I made a chart that highlights 6 specific sections of lab notebooks, each of those sections splitting up into various sub sections.  The arrows represent the flow from one section to another.  For example, within your PLN you will have experiments.  These experiments will have protocols and procedures.  Now, coming back in, say you have an idea for an experiment.  You would create a protocol and procedure for that experiment, which would ultimately go into the PLN.  After making this graphic, it was pretty apparent that PLNs lacked a communication piece, as is illustrated by the blank section after communication.

Slide06

Now, if we display the ELN properties in the same graphic, you’ll notice that the communication section starts to become filled in.  The new white sections with grey text represent the potential attributes that ELNs could have in the future.  This successfully illustrates the importance of ELNs for research communication, which is extremely important to not only present the work you’ve done, but also collaborate for future and current work.

Slide07

Here is a side-by-side comparison of some additional differences between PLNs and ELNs.  First, and most obvious, is that PLNs require you to handwrite most of your experiments, which could be time consuming and frustrating if a minor change needs to be made.  ELNs, on the other hand, retain all of your data digitally, making it easy to not only add data, but also make changes if necessary.  Because the data is stored digitally, this can make the notebook much more organized and allows you to separate your experiments easily, instead of trying to keep track of each experiment and which PLN/page number it is in.  Finally, as mentioned earlier, the communication aspect of ELNs make it appealing in terms of time taken to submit/share data.  Imagine if there was some sort of social networking site for researchers where they could publish any result directly from their ELN and get feedback.  This is one thing that isn’t possible with PLNs, resulting in more time spent actually asking for advice or publishing data as opposed to continuing research.

Slide08

Here is a screenshot of one of the ELNs we chose.  Most of the other ELNs were set up like this, where there are widgets on the side that can be dragged into a continuous document.  You can see here, I have a body text widget with some materials and the start of the procedures, and below that I have an Excel worksheet.  The widgets make it very convenient to add new things or go back and change new things that you wouldn’t be able to do in a PLN.  For example, if I make a mistake in analyzing my Excel data, I can simply change it directly from within the ELN, whereas in a PLN, I would have to open up Excel on the computer, change the data, print it out, and replace the old data by cutting and pasting the new data on top of it.  This saves an incredible amount of time in the long run and makes it easy to ensure that all of the correct and necessary data is present in the ELN.

Slide09

So what did I learn from this project?  First off, I learned about what types of data would go into a lab notebook, from graphs, to calculations, to materials, as this could have an impact on which ELN a professor would decide to use.  Second, I was surprised at the lack of ELN options.  Coming into this, I assumed there was some sort of iPad/stylus interface that could be written in like a regular lab notebook, but also had some of the widget capabilities explained earlier.  However, I found that most ELNs are web-based programs that need to be run on a computer and don’t quite have the portability I was expecting.  Finally, I learned about how any lab notebook can have a great impact on the lab, and it’s important that it is not only easy to maintain, but can be done in a way that will be able to support the experiments performed.

Slide10

While ELNs still aren’t the primary use for a lab notebook, I think that there is a potential future for them in the lab.  First, I believe that at some point, most labs will adopt ELNs to take care of certain tasks, whether it’s simple things like tracking reagents/equipment, or something more intricate like keeping all of the data and making measurements.  However, one thing I think that’s holding people back from this is the complex process that is required to transition from PLNs to ELNs.  This would usually involve a company representative training the staff on how to use the notebook, as well as implementing a server to aid in cloud storage.  On top of that, if an ELN is going to be adopted university-wide, it could take months to actually get the system up and running since more people would need to be trained.  Furthermore, the process of transporting the old data from PLNs to ELNs would take time, and many researchers might not be willing to put in that much effort to make everything digital.  Finally, if companies would like to see more incorporation of ELNs, I would suggest trying to make them more portable and user friendly.  As I mentioned before, most of the ELNs now are web-based apps and can be somewhat cumbersome and slow to work with; therefore, a cross-platform, stand alone app might be a better option.

I’d be happy to take any questions anyone has.  Just post in the comments and I’ll try to get back to you!

-Zac

Choosing an ELN

Last semester I spent the majority of my time deciding on what was important in research and how that could be incorporated into the ideal ELN.  Starting this semester, my goal is to choose 3-5 ELNs that are currently on the market and have different research groups across Columbia test them out and choose which one performs to their satisfaction.  Before recruiting research groups, I need to choose ELNs that both encompass criteria allowing it to replace a PLN as well as have some sort of free trial so that I could try it out and decide if it will fit in with a research driven faculty member.  Furthermore, the ELN must have a fairly intuitive user interface for me to navigate around and should be easy to install.

Based on my research from last semester, there is a range of potential ELNs from open-source, “make it yourself” software to companies that provide a user interface, and store and protect data.  These ELNs were narrowed to a group that, based on their websites/advertising, appeared to fill the necessary requirements to replace PLNs.  These choices were: LabArchives, eCAT, iLabber, LabTrove, GitHub, WordPress/blog, Perkin Elmer, Open Atrium, and Etherpad/Google Drive.  Here’s why each is or isn’t likely to be selected for testing by faculty.

LabArchives

Because this ELN is not compatible with Mac OS, faculty members will likely not test it.  Because people tend to use different operating systems, it is important that the ELN can work for all operating systems to allow for better collaboration.  Upon trying to download a free trial, I was notified that a version for Mac OS was not yet available, but will be in the future.  Due to time constraints, this essentially eliminated this ELN because we want it to be convenient to access to all faculty members who choose to use it.  This may be a viable option in the future, should they come out with a version compatible with Mac OS.

eCAT

I am currently testing out a free trial of eCAT.  This web-based program was easy to get access to and was fairly easy to navigate around.  I have yet to test it out on my tablet to see if everything renders well enough for it to be used as a cross-platform tool.  While I do believe it could be a bit more intuitive, I think it has a lot of features to offer and could be suitable for a faculty member depending on how they plan to use it.  With that being said, this will likely be selected for faculty members to test out.

iLabber

This ELN offered a free trial; however, I was required to submit my personal information, which would be used to contact me after the trial ended to discuss purchasing and training.  Because other options are available, I decided to pass on this ELN because if I would have to give my information, all of the faculty trying it would too.  This presents a situation that essentially forces faculty members to discuss purchasing a product that they may not be interested in and would likely result in some faculty members testing other ELNs while avoiding signing up for the free trial for this ELN.  While we don’t want each individual faculty to be contacted, we may look into getting a site license at a future date.

LabTrove

After downloading the initial LabTrove files, I discovered that the program actually required the installation of other software from the command line that would only work on machines running Debian.  Assuming that most faculty members are not running Debian as their primary OS, I decided not to use this ELN.  Furthermore, the amount of time needed to install this ELN might be an issue, especially if the faculty installing the software are not familiar with prompts and navigating and installing from the command line.

GitHub

While not technically an ELN, GitHub offers a space for research faculty to share and interact with other research groups.  The main problem with GitHub is that it requires learning different code syntax in order to operate it.  Because of this, I decided not to use this as an ELN.

WordPress/blog

Similar to GitHub, WordPress or some other type of blog isn’t necessarily an ELN; however, it will provide an easy means to share data over a group of people.  One of the major flaws of a blog environment would be the lack of proper formatting when it comes to data entry.  For example, it would take more work to enter in a spreadsheet worth of data, or type out a certain equation.  Because this would probably be more time consuming than preparing a PLN, I decided to forgo this option.

Perkin Elmer

The ELN Perkin Elmer offers has so far been the easiest to use for me while also offering options such as procedure templates to make research more streamlined.  Because the interface is web-based, it also worked on my tablet but wasn’t the most friendly environment in terms of font size and selecting different options.  Despite this, the fact that I was still able to view and edit experiments on my tablet makes this option extremely portable and allows for potentially using it directly at or away from the lab bench.  This ELN will probably be selected for faculty members to test.

Open Atrium

This is another open source platform that works with Drupal to allow users to create and manage their own website.  The goal of Open Atrium is to provide an environment that makes it easy to build a collaborative website with features such as drag and drop widgets.  It also claims to work on a variety of systems, making it easy to transfer from doing work on a mac or pc to a tablet.  The biggest downside to an ELN like this is it’s required knowledge of a new software platform.  Furthermore, the fact that it’s open source means that developers could ditch a project that a faculty member uses frequently.  This can be frustrating and potentially detrimental to the ELN, therefore, this environment will not be used.

Etherpad/Google Drive

A benefit of Google Drive or other collaboration tools such as Etherpad is that they offer real time viewing of people making changes to documents.  This could be helpful if a protocol or procedure needs to be revised by a lab.  Also, because Google Drive will store this document for you, it is possible to share all documents within the lab by creating a single Google account for the lab that everyone has access to.  While this method is not exactly defined as an ELN, it does help with collaboration and storing all documents in one place.  The biggest downside to something like this is probably the lack of digital signatures to prove from an observer that the data was collected ethically.

Once I have chosen the ELNs, I plan to disperse them to faculty members who are interested in making the switch or would just like to help test the ELNs.  The faculty will have a certain amount of time to test and comment on each ELN about their performance and what they liked and didn’t like.  I am also developing a rubric that faculty members will fill out to evaluate each ELN.  Once each faculty member has evaluated all the ELNs, further steps will need to be taken to see if the chosen ELN will be satisfactory for all faculty and plans will be put in place to incorporate it into as many labs as possible.

ELN Incorporation Into Research

Over the course of about 2 months, I have been researching not only currently available electronic lab notebooks (ELN’s), but comparing and contrasting the characteristics of these notebooks to learn more about how they will fit in to the overall flow of an experimental design.  My first step in finding an ideal ELN was to perform a simple Google search and find what companies are out there that offer this service.  To my surprise, many of the companies in question don’t offer what I was expecting: a simple replacement of the paper lab notebook with something like an iPad or Android tablet and app.  Despite this, I continued my search and found about 10 options that could at least be a start in transitioning from paper to completely digital.  Most of these options seem to focus on the topics of collaboration, creation of protocol templates, and audit trail establishment.  Two examples of the ELN’s I found are labarchives, and eCAT.  From that point, I decided to review the options more closely and make a table comparing them in the following 7 categories: security, storage space, upload limit, price, protocol template availability, collaboration, and producible audit trail.  This revealed the pros and cons of each system and revealed that there wasn’t one single ELN that immediately jumped out and was leaps and bounds better than any other ELN.  An example of the table is shown below highlighting the features of labarchives and eCAT.

Example of Table Highlighting Features of ELN’s

As it became evident that no one solution would satisfy all of a researcher’s needs, questions began to come up about the experimental process and how ELN’s compared to standard paper lab notebooks throughout the research process.  This prompted me to develop several flow charts that outline the roles of both ELN’s and paper lab notebooks in the research workflow to graphically show what properties of ELN’s are necessary to make it a viable and worthwhile option.  Furthermore, I added options that I think will be important in the future ELN’s (many of which are in the “communication” category) and greyed out others to show where the technology may be headed and to show the potential power of ELN’s if they are to be universally adopted.  I made two additional flow charts to illustrate both the data collection and data retrieval process and asked several questions about storing and harvesting data, while comparing the effects that either a paper lab notebook or ELN would have on the specific questions.  All flowcharts can be seen below.

Paper Lab Notebook in Research Process

ELN in Research Process

Data Storage

Data Retrieval

The work done so far tries to lay a landscape for the benefits of having an ELN over a paper lab notebook.  Going forward, this information will be useful for weighing the pros and cons of incorporating ELN’s into the Columbia faculty’s labs.  Towards the end of the semester we hope to line up at least a couple of the companies for trial licenses that professors will be able to utilize to see what works for them.  Once we have a general idea of what the needs of the Columbia community are, we can hopefully start integrating some of the main highlights of ELN’s and begin the transition from paper to digital.