Laboratory notebooks: paper tradition or digital future
- The turning point in the 21st century
- Changing practices
- Choosing an electronic laboratory notebook
Some have high heritage value such as Pierre and Marie Curie’s laboratory notebooks kept and digitized by the French library, Bibliothèque Nationale de France. Others are the subject of incomprehensible legal battles (1) in a race for innovation and filing patent claims. Finally, others “sleep” in laboratories waiting to be consulted one day.
An essential scientific accessory, the purpose of a laboratory notebook is to faithfully reflect the researcher’s work. Researchers note information concerning the operating procedure of their experiments in detail: date, goal and experiment number, bibliographic references, weight and pressure measurements, heating time, description of reaction scheme, observations, reference to raw data stored electronically, illustration of results with graphs glued into the laboratory notebook, etc. This rigorous exercise might seem banal, but it is crucial to ensure that the experiment can be reproduced, which is the only way to validate results.
The turning point in the 21st century
In around 2005, the Ministry of Higher Education, in partnership with INPI, began to recommend the use of standardized laboratory notebooks. Some 100,000 notebooks were distributed free-of-charge in 2006. Unprecedented, they featured a page numbering system and a text box for a signature and the date of each experiment. During their thesis, a PhD student may fill up to 10 laboratory notebooks. That makes thousands of laboratory notebooks to archive for all the groups in just one laboratory. Theoretically, maintaining these records is subject to guidelines to guarantee scientific integrity :
- Write in indelible ink.
- Do not tear out pages, do not whiteout but score out.
- Write a summary.
- Authenticate documents glued in with a signature that overlaps the page of the laboratory notebook.
- Use close writing so that no other information can be inserted in the text.
- Always sign and have the laboratory notebook counter-signed, etc.
Observations in situ in laboratories tend to describe a slightly different reality. « What is written down […] is often only legible by the researcher alone, who glued in, for instance, spectrophotometry results, scribbled formulas and diagrams, summarized the steps of a chemical synthesis sometimes forgetting certain details, sometimes forgetting to enter them at all, particularly when the experiment takes less than ten minutes, etc. In short, laboratory notebooks are neither exhaustive nor completely readable by a person unfamiliar with the operations (2). »
These observations were confirmed by interviews with PhD students as part of the Datacc’ project.
Searching through a colleague’s lab notebook can be really difficult because it’s poorly written or badly summarized. Sometimes it’s like reading a novel. In others, the English was very approximate.A Chemistry PhD student
Best practices being used in a laboratory notebook depend on each person’s will and discretion which prevail over standardized guidelines: « There are no laws or regulations, not even a memo that provides guidelines for contents of the national laboratory notebook, their use in research departments, let alone how rules for use should be observed. Given the insufficiently identified supports, their legal value is also very questionable (3). »
This informal approach contrasts with the supposed documentary nature of laboratory notebooks. It is the laboratory notebook that proves the paternity of a result (the identity of the researcher at the origin of the discovery) and its priority right (date of the discovery) in a court of law or when filing a patent.
In theory, the laboratory notebook has an intrinsically high value: it is a documentary tool for scientific reproduction and legal protection. It is questionable to continue to entrust such evidence to a degradable paper medium, given the risk of incidents (floods, splashes, etc.) and the difficulties in sharing and reusing, even within the same group. Now that projects are often conducted between groups from different institutions and laboratories, practices using digital tools instead of paper laboratory notebooks are likely to spark a strong evolution of practices:
- A networking approach between researchers (which can be adjusted through administrative rights)
- Better traceability of content with time-stamping and mandatory fields for experiment descriptions and electronic signatures (4).
- Automation of some tasks considered tedious such as experiment reproduction.
- Better connection between experimental specifications (traditionally described on paper) and results (digitally native). (5)
The use of electronic laboratory notebooks has boomed over the last few years. Nowadays, there are reportedly more than 100 software products on the market (6).
The most complete and prominent experiment in France was conducted at Inserm with widespread use of Labguru in all the institution’s laboratories, including those under multiple authorities.
Other institutions chose to use open-source solutions such as eLabFTW, used by Institut Curie laboratories which enouraged the tool’s development. To a lesser extent, elabFTW is used in some departments at Lyon 1 University (for example, LBBE in biology, the biophysics group of ILM). “In-house” solutions are sometimes tailored to meet the needs of certain groups. This is the case in Lyon 1 University , where a group of chemists from the ICBMS uses i-labchem, available under Mac. For over ten years, on the LMA platform of the Institute of Physics of the 2 Infinities, physicists have been using OpenSourceLogbook to develop Virgo for gravitational waves as part of a European project. In Grenoble, the software team of the PaNOSC project (7), award winner of the H2020 program, working with Synchrotron, has developed the elogbook solution. The tool is closely linked to the design of the data management plan chosen by project participants. « Experiment methods and the scientists’ comments are necessary for understanding an experiment performed on a beamline. This is considered to be required in the metadata that will be published with open data (8). »
After several months of using Labguru, a project engineer in a team from Inserm made three observations:
Experiments are much more reproducible. We weren’t necessarily expecting that. The number of useless experiments is greatly reduced, and I don’t waste nearly as much time trying to reproduce experiments.A Biology Project Engineer
The guidelines for the data management plan proposed by the ANR rightly refers to “laboratory notebooks” as a data documentation tool, but it does not specify whether it should be paperless or digital. Yet the transfer from paper laboratory notebooks to electronic generates a fragility of the data’s traceability.
On a different note, the electronic laboratory notebook inspired an interesting educational experiment conducted amongst students of Université de Grenoble Alpes and high-school students in several Rhone-Alpes schools. Baptized LabNbook, the tool is designed for use in class.
Choosing an electronic laboratory notebook.
General or specific to your discipline?
There are two types of tools on the market. The first are so-called generic laboratory notebooks, adaptable to several disciplines (chemistry, biology, physics, etc.) and different contexts (laboratory, start-ups, etc.). In this case, the aim is a user-friendly tool that is easy to use (e.g. elabFTW). Some tools even allow plug-ins for particular disciplines (e.g. LabCollector and its Chemistry plug-in).
To help you recognize the difference between the available tools, we have classified six solutions, most of which are designed for chemists but not only.
The second type of tool is specific to a discipline. In this case, functions have been developed to work with applications, databases and instruments specific to a discipline. This is the case for Mbook (interfacing with Mnova for the analysis of spectra, stoichiometric calculations) and eNovalys, interfaced with PubChem. Discipline-specific tools have the ability to search by chemical structure and sub-structure within the same electronic laboratory notebook, saving researchers’ time.
What about security?
Data security is a recurrent concern for researchers. The degree of confidentiality required by their work varies depending on the commercial outcome of the project and industry partnerships.
Most available software offers an SaaS (SoftWare as a Service) type formula which frees laboratories of hosting constraints and tool maintenance. The supplier “takes care of everything”. On the other hand, this raises questions about data confidentiality stored externally on servers or in clouds that are not managed by the institution.
Laboratories are increasingly situated in restricted zones [zone à régime restrictif (ZRR)] where sub-contracting hosting to software publishers is not authorized. The only viable option seems to be local installation. This constraint may hinder the widespread use of electronic laboratory notebooks in laboratories that do not have the necessary technical assistance.
Storage management varies greatly from one laboratory to another. It can be very informal, dependent on the memory of the members of the laboratory. An Excel table can be used, albeit with limited capacities when treating a large number of substances (e.g. chemical libraries). Or specialized tools such as Chimie Tech are available.
Apart from the experiment management function, most electronic laboratory notebooks can also directly manage stocks of products, apparatuses and lab samples, through a solution called Laboratory Information Management System (LIMS). This provides centralized access for inventories, ordering products and ensuring that stocks of a substance are available for an experiment. These tools can usually be imported to Excel type files for automatic entry of products in the database. Files must, however, be configured according to the tool’s instructions.
At what cost?
Generally expensive, the cost of licenses is calculated according to the number of users, so it soars if the tool is made available to hundreds of researchers. In most cases, the price is between €100 and €150 per user and per year, and this does not include technical support. Perpetual licenses can also proposed (e.g. €3,000 for 5 users for LabCollector). Some solutions are free for universities, but they offer less developed features (e.g. no electronic signature). This is the case for FindMolecule, as indicated in the summary table here.
Converting to a digital solution is a significant change in method, but it is does not necessarily mean abandoning paper completely. Hybrid practices are emerging where, for personal convenience, researchers continue to write down their measurements or other observations in their own paper notebook, before entering all the information in the electronic laboratory notebook. Others prefer to centralize all the information and use their electronic laboratory notebook on a tablet or computer for instance, while conducting experiments.
Finally, careful consideration is necessary when making the choice between paper and digital tools. Three issues influence the decision: cost (paper is much cheaper, even if the non-traceability of experiments represents a cost), use (getting accustomed to a tool can take time, paper is handy on the lab bench) and environment (data security issue).