ELN - Electronic Lab Notebooks
ELN - Electronic Lab Notebooks
The opportunity to switch to electronic lab notebooks (ELNs) had been discussed for 2 or perhaps 3 decades without any concrete change. However, the recent boom for tablets, ipads and the new amazon fire costing less than one pipette, is pushing toward a ELN resurrence. We will have ELNs in the lab soon? In the last months, I gave my advisory counseling to some companies considering to enter this market, here I summarize my thoughts about ELNs.
Is an Electronic Lab Notebook right for you? Free whitepaper, dowload now.
For my experience, speaking with senior PIs, the most frequent points against ELNs are:
1) Doubts about patent application eligibility of data recorded electronically (cheating is easier on files than paper, think to edit back a file date vs rewrite a fake classic lab-book with old ink on old paper).
2) Fear to loose the data: files get more easily corrupted than paper and senior PIs are still thinking to backups with the 'floppy-disk' in mind.
My point against current ELN is on the longevity of the file format the data is stored on. In most cases, getting into the market implies inventing a proprietary file format. As it turns out, I have problems to open files created 10-15 years ago because the software is dismissed or not compatible with new operating systems, but I can read a lab book written by my PI 30 year ago. Open formats are preferable to assure that the data will be safe also if the company/product will exit the market.
My positive point is that I can not have such a great benefit in reading an old lab-book because I don't understand the calligraphy or because some points are omitted or spread on different pages/different books. Of course, a search function is not available!
What is the principal ELN function?
In my opinion, the first aim of a lab book is to guarantee the repeatability of the experiment and to summarize the most important results. Summarizing results it is easy on paper: just need to attach a graph or a picture with papertape... But the repeatability is a pain! You have to manually write each detail of the protocol: the batch of the reagent, the concentration, the volume, the number of times you mixed it, the number of times you thawed it. The amount of time you left the tube on the instrument/bench/freezer... You can easily spend a day in writing the details, but Darwin times are over and you are in hurry! Getting your work published today requires 10x more figure panels than in the past, conversely the time to graduate or to get-tenure is the same or less! Today, research scientists are desperately running, so it is natural and unfortunate to evolve a behaviour that skips the parts that are not essential for publication (i.e., keeping a fair notebook). However, the repeatability is then compromised, and the retraction rates are increasing.
What ELNs should do?
An ideal ELN would be so intelligent to auto-fill the details I'm lazy to write down, like a personal assistant that just follows you at the bench and keep notes of your handling without you need to stop. The ideal personal assistant, will have also some memory: ELN, when I used this polymerase last time? It was two weeks ago for the experiment Y. I mean, the ELN should be interrogable, like google is.
The biggest mistake ever done in ELN design
Do not design an ELN as a closed environment that manages lab data. Every lab, every scientists has their own preferences in terms of storing some report on some software: the same graph can be made in excel, open 'excel' alternatives, in graphpad prism, copied to word, copied to powerpoint, distilled to pdf, to jpg etcetera. Scientists are already asked to change working hypothesis/protocols different times a day, dont ask them to change also the 'system' they store the data, they need some sticky habits.
ELN is a personal secretary, nothing more
A good ELN, as a personal secretary, should just be flexible and smart enough to point/link to the different files with a more elastic mind than windows explorer or mac. For instance, I can store my qPCR files in a qPCR folder, this is beautiful if I then ask how many PCR I have done this year, but ugly if I need to retrieve 4 PCRs for the manuscript Y or the 3 PCR paid with a grant X. I can store the PCR for the project Y in the Y folder or the X PCR in the grant X folder. But I'll rapidly mess up stuff. As a clear example, in the last 5 years, my 'Results' folder contains 1905 sub-folders stuffed with 22,187 files for a total of 22 Gigabytes, and I'm not involved directly in any massive sequencing, the core is all about basic western blots, luciferase assays, RT-qPCRs and some microscope picture!
My ideal ELN would be able to assign my preferred file (protocol, raw data file, result file, literature reference, manuscript in preparation, presentation, referee,...) to each of their corresponding folders (project, person, experiment, grant, application), and perhaps keep an historical protection of different file versions. In my mind, more than a software, the ELN should be a sub-operating system able to run as a virtual environment into win, mac, linux and able to mangle my data better than I do, like a personal assistant.
If you are developing some ELN stuff and want my feedback, you can get in touch at firstname.lastname@example.org, confidentiality is guaranteed.