Depending on the scope and complexity of the research at hand, the items of equipment needed to perform vital tests can be numerous, sizeable, and expensive. The lack of available lab space, investment funds, and even time needed to conduct tests on these advanced machines can make it difficult (if not impossible) for start-up and mid-size operations to get their hands on such machines, and thereby compete with bigger contract research organizations (CROs).
Emerald Cloud Lab’s novel solution enables even humble-sized scientific operations to make use of high-end laboratory equipment by offering remote access to such advanced gear remotely, for a monthly fee that amounts to less than the cost of investing in a single piece of equipment. Toby Blackburn, ECL’s vice president of business and development, took the time to speak with Outsourcing-Pharma about the advantages the solution offers, and some of the notable organizations that have taken advantage.
OSP: Could you please talk about the concept of “automated science” and how researchers are growing to understand and accept it?
TB: Researchers in science are beginning to look around at other industries (tech, manufacturing, distribution, etc.), and the advancements that technology has made in those areas, and are asking themselves why they are still operating labs the same way they did more than two decades ago. The technology required to bring the same efficiencies, cost savings, and process improvement to the lab environment is well developed and proven in other industries.
The challenge for the research community has been the degree to which high-skill activities (experimental design, data interpretation, application to molecule/process) have been intertwined with low-skill activities (procurement, inventory management, data recording, sample, and material logistics). The time needed to run a lab is taking time away from research and discovery.
Cloud labs create a separation between scientific discovery and laboratory management, allowing your scientific talent to focus on the science. To realize this vision, our software platform uses AI to intelligently suggest experimental parameters based on sample and other associated information, making it easy to ensure that experiments are fully and unambiguously specified from start to finish.
Fully proceduralized and defined methods are a by-product resulting from day-to-day research, allowing for directly reproducible experiments without additional effort.
Cloud labs also reduce costs by removing the need to construct lab space, purchase equipment, and hire staff. Finally, since the cloud lab runs semi-autonomously at all times day and night, researchers are advancing their experiments much more quickly than if they had to depend on their own collection of instruments.
There is a whole host of positive externalities that arise from this "proceduralization" – remote execution, the ability to move up a level of abstraction (that is, writing a protocol that automatically generates a method, versus writing a protocol that runs a method), cost efficiencies, increased throughput, model quality data on the output, etc.
Functionally, all of Emerald Cloud Lab’s value propositions come out of this fundamental desire to fully define science, using the cloud as the interface between the scientist and the lab.
OSP: The ECL is a novel concept—could you explain how the idea came about? What needs did you and your team see in the industry you felt this solution could address?
TB: Emerald Cloud Lab arose from a need that our co-founders, Brian Frezza and D.J. Kleinbaum, had with their first startup, Emerald Therapeutics. Emerald Therapeutics had a research plan that was much larger than their budget, which necessitated working extremely efficiently. One way they achieved this was to train each scientist to write code so that they could automate aspects of their work whenever possible, thereby multiplying their productivity.
Of course, scientists use the results of one experiment to inform the next experiment, and it quickly became apparent that a bottleneck in this environment was the ability to seamlessly share data between scientists and their experiments, so the founders wrote a unified data-sharing platform for the company.
Having successfully unified and automated experimental data analysis, they then asked themselves if they could go a step further and automate the generation of that data. To make this a worthwhile endeavor such a system should make it easy to specify and run any method that an Emerald scientist wanted, otherwise the time spent programming the method would exceed the time to just run it by hand, which would defeat the point.
Solving this design brief led to the precursor of Emerald Cloud Lab, a unified software environment that allowed scientists to quickly script an experiment and analyze the resulting data. After some time in development, the productivity gains exceeded everyone’s expectations and the founders and board realized that Emerald Cloud Lab could enable other startups and large companies to similarly leverage their scientific talent in ways not possible before.
OSP: How does it work?
TB: It’s easiest to understand how Emerald Cloud Lab works by breaking the process down into four steps: Command, Run, Explore, and Analyze.
In the Command phase, clients ship their samples to an ECL facility and design their experiments in the ECL Command Center application. Every detail of every experiment can be designed without any artificial restrictions — it’s the same as being in the laboratory itself.
ECL can handle materials as small as microliters and micrograms or as large as liters and kilograms, and experiments span the range of capabilities needed in a modern pharmaceutical company. With ECL’s Command Center software, any experimental parameters can be set or modified with a point-and-click interface – or even by directly typing commands into your lab notebook. Once designed, protocols are automatically saved to a database for push-button reproducibility and higher-level scriptability in larger workflows.
In the Run phase, ECL remotely conducts the experiments in a highly automated facility to the scientist’s exact specifications. The lab currently has over 190 different types of instruments online, including HPLC, Mass Spectrometry, and NMR. Once an experiment is submitted through ECL Command Center, it is run automatically. Results are added to the user’s database in just a few days.
In the Explore phase, clients can make use of the “ECL Constellation,” a network of linked database objects that structures data in a highly organized knowledge graph. This graphical database grows automatically over time as more experiments are conducted, containing all data associated with every experiment ever run on the system in an easily searchable format.
We really mean all data here - whether you want to look up the model number of a column used in an experiment, or the lead time on a specific buffer, or the number of minutes a particular sample spent outside of the freezer during testing. Because ECL Constellation lives in the cloud, data is accessible from any computer with a secure login.
With ECL Constellation, any questions about experiments can be answered in seconds by surfing through this knowledge graph with a few clicks or keystrokes. Searches can be conducted across the full history of experiments run on the system, including system data like the QC reports that ECL runs regularly on its instrumentation.
In the Analyze phase, researchers can use the ECL Command Center for access to an extensive suite of tools to plot, analyze, and visualize results. ECL Command Center offers over 4,500 powerful functions for data visualization, analysis, and simulation.
The software also allows experiments, data, analysis, results, and scientific figures to be exported, shared, or published on the web. These tools can be accessed through a point-and-click interface, or the commands can be directly entered into your lab notebook. This makes it easy to repeat or scale any analysis with a single command and to automate report generation through higher-level scripting.
OSP: Could you please talk about the benefits users of the cloud lab might gain? Please feel free to talk about anything from time savings, cost savings at various levels, the ability to ‘borrow’ equipment rather than invest, etc.
TB: Our clients’ proof of concept results have demonstrated five- to eight-time improvement in individual scientist productivity, resulting from both the reduction in logistical tasks required of scientists as well as increased instrument utilization from the ECL operating 24/7/365. This translates into cost savings and increased throughput.
The data that comes out of the cloud lab is more organized and complete than anything on the market. We call this "model quality data," in that it is fully ready for artificial intelligence and machine learning tools without the type of data cleansing required today.
The data is also much easier to access. With the traditional model, a glitch in a manufacturing process could mean dozens of scientists searching for, aggregating, and organizing relevant development data sets for weeks, just to be able to start to interpret the data and make informed decisions on what to do next. By contrast, with Emerald Cloud Lab, the same data could be identified and organized in a matter of minutes.
For startups, the breadth of equipment available without incurring significant capital costs is a huge driver. Emerald Cloud Lab is based on an annual subscription model, and startups can obtain a full year of access to the system for about the cost of a single piece of laboratory equipment. Clients can scale up their use of the lab without taking on additional expenses--not only in capital but also in human resources.
Additionally, clients can get started in days, instead of the many months it would take to find, build out, purchase, install, and run equipment, etc.
OSP: How does the solution differ in what it offers clients compared to on-demand lab space?
TB: First, ECL’s solution is comprehensive; for a single access fee, clients get access to over 190 types of lab instruments (and over 550 individual instruments), full-featured lab notebooks, automated inventory management, and access to a full catalog of reagents, consumables, and other materials. These instruments are always returned to "ground state" and routinely monitored with quality control experiments to make sure they are always ready to run a client's experiments.
Overall, with ECL, researchers can run all of their basic experiments exactly as they envisioned without the hassle of maintaining their own laboratory.
Furthermore, on-demand lab space still requires capital investment and buildout time – which can be six months or more for lab fitout, instrument acquisition time, and so on. By contrast, a researcher can start experimenting with ECL in days.
Even the most well-equipped labs still lack much of the equipment that would be core to biotech/pharma research. Additionally, shared equipment is exactly that – shared. Without the data records and attention to detail in returning an instrument to “ground state” that ECL provides, your confidence in your results can be really shaky if you are always worried about the possibility of an instrument problem.
OSP: Could you please talk about some of the use cases and customers the ECL is best suited for? You mentioned small-scale and start-up operations, for example.
TB: Emerald Cloud Lab is well-suited to companies of any size that are interested in improving the means by which they approach life science research. Our use cases range from top 20 biotech companies to fast-growing startup operations to Carnegie Mellon University’s Mellon College of Science.
Certainly, small-scale and start-up operations can immediately benefit from the cost savings and risk reduction that Emerald Cloud Lab offers. For the cost of a single piece of laboratory equipment, these companies have access to over 190 different types of high-throughput instruments.
Emerald Cloud Lab saves startups from having to raise money to purchase and maintain scientific instruments. It enables greater repeatability because all protocols and results are recorded, tracked, and followed automatically. It also frees up crucial time needed for experimental design and data interpretation.
Of course, those benefits apply not only to fast-growing start-ups but to larger biotech companies and academic research labs, as well. In fact, one of our highest-profile biotech clients conservatively estimated that their scientists saw productivity improvements of 500-800% with Emerald Cloud Lab over a traditional lab. Separately, a Carnegie Mellon graduate student was able to replicate four years of dissertation research in two weeks with ECL.
OSP: You’d mentioned some CROs might view the ECL tech and capabilities as a threat to their operations—could you please talk about that?
TB: It’s true that some CROs might view Emerald Cloud Lab as competitive to their business. However, these organizations can benefit from the scalability and risk reduction offered by Emerald Cloud Lab in the same way any pharma client might. Access to ECL will enable them to compete with much larger players.
On another level, to apply their scientists’ expertise, CROs have been shunted into the role of equipment operator. We see many synergies in how we are able to leverage the expertise a CRO provides with the highly leveraged operation and advanced data processes that a cloud lab enables. We strive to ensure equipment operates as intended, and experiments are performed as specified.
As we discussed, researchers around the world, including people at Carnegie Mellon University, found their work interrupted by the pandemic, being unable to visit their physical lab space. Could you please talk a bit about your collaboration with CMU—how you got to know each other, those early talks, the coursework, etc.?
Emerald Cloud Lab’s founders, Brian Frezza and D.J. Kleinbaum, are both alumni of Carnegie Mellon University’s Mellon College of Science. ECL’s collaboration with Carnegie Mellon University grew out of a desire of the founders to give back to their alma mater – along with a challenge that was issued to the Dean of the College of Science, Rebecca Doerge, when she arrived at CMU some five years ago.
At the time, Dean Doerge was given the goal of elevating the reputation of the Mellon College of Science to a status comparable to the university’s Engineering and IT programs and decided that partnering with ECL was the best way to fulfill this objective.
The burgeoning relationship proved its value when COVID-19 forced many colleges to shutter their labs – threatening to derail or delay their research for months or longer. Emerald Cloud Lab donated their time and their cloud lab infrastructure to Mellon College of Science researchers to enable research to continue despite pandemic restrictions.
This relationship allowed CMU undergrads an unprecedented opportunity to learn about automated science. Students spent six weeks studying the interactions between small-molecule dyes and proteins, running their experiments in the cloud.
This area of research, the modeling of ligand-receptor interactions, is fundamental to many pharmaceutical and biomedical research objectives. It is potentially related to both biotechnology and biofuel advancements as well.
Beyond bringing research back to shuttered labs, Dean Doerge sees the advent of cloud labs as the future of scientific research. To that end, Carnegie Mellon is making a $40m [USD] investment in the college to construct their own version of Emerald Cloud Lab’s facility as the linchpin of a multi-year strategic initiative.
OSP: Carnegie Mellon professor Marcel Bruchez has said that this technology stands to make science more “inclusive.” Could you talk about how that works, and why that’s a brilliant thing?
TB: The notion of inclusivity and democratization is at the core of what makes cloud labs in general – and Emerald Cloud Lab in particular – such an important development in life science research. Promising scientific minds will no longer be constrained by the realities of laboratory access, where only the largest, best-funded organizations in the wealthiest countries can make important, sometimes life-changing discoveries, and scientists must be physically present to play a role.
The cloud lab enables biotech start-ups to get up and running at a fraction of the cost of what it may have taken only a few years ago. With Emerald Cloud Lab in particular, more than 40 types of important biotech experiments can be conducted while the scientists are sitting in front of their computer – wherever that computer may be. Automation eliminates unintentional bias, and reproducibility allows scientists to collaborate and cooperate across the vast expanse of the globe.
This inclusivity and democratization will undoubtedly lead to greater scientific advancements faster, as the concept gains in popularity among more companies and universities around the world.
OSP: Could you tell us about ECL’s expansion plans, both what you’ve got on deck and your long-term goals to grow?
TB: ECL continues to expand its suite of scientific instruments and scale our total operational capacity. When we started the company nine years ago, we made a list of every instrument and capability we wanted to include in the lab, and we are about 85% there. Along the way, we’ve invested heavily in scalability and we’re well-positioned to grow our footprint to meet customer demand in the coming years.
Also, we are excited to see Carnegie Mellon’s cloud lab open up in the next year, which will enable their students around the world to learn about automated science and their professors to produce groundbreaking research that will elevate Carnegie Mellon’s College of Science and advance life science research for the world. We believe that this is a seminal moment in the adoption of cloud lab technologies across academia and will further accelerate the adoption of cloud labs in the industry.