April 20, 2021|Cloud Laboratories, ELN, Science and Technology, Scientific Informatics
20/15 Visioneers, Leaders in Science and Technology An Industry Perspective Contents
1. History and Introduction...............................................................................
2. Observed Challenges ...................................................................................
3. Fundamental Need........................................................................................
4. ELN Replacement Strategies .........................................................................
5. The Next Generation ELNs.............................................................................
6. The Brain and Spinal Cord of Your Research Environment ............................... “By Changing Current R&D Culture and Approaches, the Convergence of Science, Technology, and Human intuition will drive unparalleled innovation.” John F. Conway 1. History and Introduction
The conceptual history of the ELN probably goes back to daydreams in the mid-1700s by the world’s more creative and imaginative scientists; however, it wasn’t until the late 1980s and early 1990’s that serious workgroups and consortiums were being formed to take it from concept to reality. This LIMS Wiki outlines the details we know of https://www.limswiki.org/index.php/Electronic_laboratory_notebook. Whether it is on paper or in electronic format, the main purpose of a laboratory notebook is to capture the scientific method. (See Figure 1.) Recording your Ideas/Hypothesis, Methods/Materials, Experiments, Results, Analyses, and Conclusions is a foundational requirement of good science. The captured details are intended to lead to enhanced scientific reproducibility and replication, (creating knowledge and enabling technical transfer), a major issue today in both academia and industry for a multitude of reasons that we may only touch on. The inspiring aspect of this is realizing the dreams of an electronic environment that allows for enterprise searching, assimilation, and curation of scientific data and business processes. After 30 years of functional prototypes and enterprise, scientifically aware capabilities and functionality, today’s modern ELNs are hitting the mark.
Figure 1 2. Observed Challenges
It has taken 30 years of ELN evolution (along with R&D evolution) to get to the current state of scientifically aware ELNs. Effective technology is here, and the capable ELNs can consume data dictionaries, taxonomies, and ontologies and effectively manage the data. They are all coming up to speed on the overarching need to incorporate detailed process handling. A great deal of work remains to incorporate what is known as FAIR data standards into the curation and distribution of experimental data. This lack of process detail and accessibility has led to a multiplicity of incompatible software solutions. Unless FAIR data standards are incorporated into the ELN an organization risks ending up with multiple workstreams and solutions. This can cause an erosion of data integrity that ends in non-compliance and serious gaps in Figure 1. So the excuse or offering that we are not a LIMS and we are not an ELN may be a minimal viable product that either has to become an ELN or a LIMS or better yet a hybrid of the two. The biggest challenge that remains is the culture of adoption and use of an ELN. Many of the compliance issues, though not all, are excuses, and luckily, we have the drug development areas to point to where rigor has been mandated and regulated. Non-compliance in these groups is mostly a non-issue. After 30 years of ELN requirements gathering, business analysis, and implementations the team at 20/15 Visioneer's has heard all the excuses and reasons research groups can’t adapt to and adopt an ELN. Here are some:
Table 1 3. Fundamental Needs
Any organization deploying an ELN must thoroughly consider the end-user issues, namely, usability, user efficiency, flexibility (where needed), appropriateness, and the major selling point - WIIFM (What’s In It For Me). As a basic analogy, would you expect your automotive mechanic to repair your car with a Swiss Army Knife? That is what we thought. Would you expect your diversity of scientists to use an ELN that was designed almost exclusively for IP capture? No, we’ve seen this many times and it’s never worked well. A well-designed ELN needs to handle natively or through integration these following important high-level workflows: Request, Sample, Experiment, Test, Analysis, and Reporting. It’s often true that some of these workflows are handled by many existing LIMs environments, especially Request, Sample, and Test. This has led to an ever-increasing overlap in ELN and LIMS, however, historically ELNs are used more in the dynamic experiment part of R&D, and LIMS are used primarily in the more static parts. Many people and organizations learned painful lessons by selecting the wrong scientific software in these environments. Finally, where a platform has both a scientifically aware ELN and LIMS, they are now solving some of these issues organizations have had to deal with for years. As mentioned earlier, the fundamental need for R&D organizations is to capture each laboratory scientists' scientific method. This doesn't mean that data scientists and computational chemists and biologists are off the hook! (There is a separate computationally based electronic notebook for you) All personas count and need to capture their experimentation, whether physical or electronic lab-based! Another fundamental need is the sharing of data, process, and experiential knowledge. Modern organizations need to be collaboration capable and or ready. Unfortunately, it took a pandemic to expose the shortcomings of organizations that lacked this readiness. Any new ELN deployment must have FAIR data and processes, period, or risk wasting money and effort. There is also precedent in the large biopharmaceutical industry for one scientifically aware ELN, (Chemistry and Biology) in research. We are not saying that you could not have 2 ELNs, one for Chemistry and one for Biology, or one for research and one for development, and the only reason you would have separate brands could be cost or legacy or the ability of a next-generation ELN to be validated in the development environment. Importantly, get your organization ready for a much higher level of automation now that Cloud Labs have gone commercial. We believe that Cloud Lab deployment and the change management this requires will be the most significant transformative factor in future R&D landscapes. Just like in the case of cloud computing (Cloud vs On-Prem) legacy manual laboratories will disappear. While start-ups and smaller virtual companies are expected to lead the wave of early adoption, there will undoubtedly be a transition period where larger biotech and pharmaceutical companies progress through a hybrid environment, with work divided between public cloud labs, private cloud labs, and legacy facilities. 4. ELN Replacement Strategies
The lessons learned here are numerous, from one of the first biopharma commercial ELN deployments to current, multi-national biopharmaceutical companies supporting several thousand users. For the latter, we will warn you that this is a complex program that will take serious change management (including Business and IT partnership!), Due Diligence, Planning, and Risk Mitigated POCs and Pilots. (Has your organization defined POC, Pilot, Prototype, and Minimal Viable Product? Don’t proceed without it (maybe move to fundamental) ) The cadence of events, like which groups, sampling, and order are all critical understandings. Change management, 80% of the time, is either done improperly or not at all leading to program and project failure. You can do everything “right” up front and initial implementation may go great, but if adoption lags it can throw an organization into a death spiral resulting in negative ROIs and 10’s of millions of wasted dollars. This doesn’t have to happen if you approach your complex situation with pragmatism and proper CHANGE MANAGEMENT! Lastly, every good strategy has an exit strategy, it should be clear and clean, and uncompromising, or it will eventually cost you significant time and effort like the current large replacements going on throughout the industry. The price tags mentioned above are serious money, and no matter the size of the organization it’s a serious capital expenditure. 5. The Next Generation ELNs
Fortunately, we already authored this article in early 2020, Visioneering the Next Generation ELN, you can access it here, https://20visioneers15.com/blog/f/visioneering-the-next-generation-eln. The Next Generation ELNs need to deliver exceptional user experiences, scalability, and scientifically aware capabilities. This last point is critical as both chemistry and biology need to be served in one ELN. Some companies will choose to have separate ones, but we strongly recommend adopting one ELN for several reasons including, IT footprint complexity, integration, cost, and lastly, the simplification of your future exit strategy, as ELN companies are going to improve on the previous enterprise quality points and upgrades and new products will be very appealing. The term What's In It For Me (WIIFM) is a critical criterion when selecting a new next-generation ELN. Remember your scientists are using this every working day in their labs, not management! Imagine using a swiss army knife as your toolbox vs proper hand and power tools! Second, industrialized R&D is now highly externalized, and having an ELN that is collaborative from both a functionality and security perspective is absolutely essential. We highly recommend selecting an ELN that was built with this in mind with Cloud compatibility, even if you have no collaborators, as this will mitigate risks and conform to the wider industry strategic direction. Lastly, plan for automation, the age of the robot is upon us and will only grow exponentially. You want your ELN to handle automation and high throughput experimentation workflows. The Cloud Lab environments are another major disruptive approach to experimentation that will play a major role in the near future for many companies. You can read more about that here: https://20visioneers15.com/blog/f/the-cloud-lab-revolution Integration with these next-generation environments is critical and another future-proofing step you can take. Remember you want this journey to last as long as possible, this expense needs to result in a nice ROI.
6. The Brain and Spinal Cord of Your Research Environment
A well-built ELN becomes the brain and spinal cord of your organization, an extension of your great scientists captured processes, data, and tacit knowledge. We think we have mentioned in the past that Sample Management is the backbone of NME discovery. Remember the fundamental reason for an ELN is to capture the totality of the scientific method. If done properly Intellectual property capture comes along for the ride. However, this is 2021, 30 years after the first rudimentary ELNs were being used at Kodak and small university labs. We now have the power of search, relatedness, unique identifiers, scientifically aware, and molecular representations that you can do substructure similarity and advanced algorithmic searches on. In other words, you should not just be able to record your scientific method electronically, but you should be able to exploit your whole connected environment to influence ideation, simulation (including prediction), design, make, test, and analyze. No more arts and crafts, you are integrated to Microsoft Office or Office-like environments, sample management (the Backbone of medium and large R&D organizations), Request Management (everything starts with a request) to an integrated reagent inventory system) and ends with Entity Registration Systems. Like a brain and spinal cord, the ELN is a fundamental systems need in a data and process-driven R&D organization. One of the most essential needs of a highly adopted and high-functioning ELN is its configured data environment. This critical step is an opportune moment to adopt data standards, process standards, and importantly harmonized and optimized processes. The latter comes at a huge expense in both cost, deployment, and adoption delays if done after the fact. Adoption, driven by usability is critical for an ELN deployment. Why? because this is a major and costly effort for most medium-sized and large-sized organizations. The coordination, the requirements, the configurations, the change management are time and effort away from some science efforts, but if done right it will prove its worth over and over. The opportunity cost of not doing it may be hidden, but can ultimately be staggering. In legacy biopharmaceutical and material science environments, data wrangling can be as high as 80%. The European Union and PricewaterhouseCoopers report on the cost of UnFAIR EU R&D data is as high as 26 Billion Euros a year. If you agree that the ELN is a fundamental and foundational approach to ensuring R&D data is FAIR, then you can easily calculate a fraction of the ROI that you get with a highly adopted New ELN in your Research environment. For simplicity's sake, the cost per EU researcher in the unFAIR data environment is up to ~15,000 Euros a year. Multiply this by the number of new ELN users and you have a significant return on investment, not to mention, you reduced the data wrangling and are able to spend a significant amount of time on more science. We have also done calculations with our clients on assessed data wrangling, and ELN compliance, and not surprisingly the ROIs are in the same ballpark as the previously mentioned calculation. Your time to market (TTM) decreases, you are preparing your organization for insilico-first or model-first approaches (secondary use of data), and you will now drive another level of efficiency by numbers like 40%! The cost of not following this approach is to risk all types of failures, including a negative ROI, because you just spent or committed to spending upwards to 30 Million dollars/Pounds/Euros and missed the mark when it came to adoption and a chance to take your organization on a successful journey, not to mention you may have just missed a critical and lucrative discovery. Everything we just discussed takes STRONG leadership. Promote the Culture, set the strategy, and reap the rewards in as little as 18 months from a FAIR data environment!