PNNL

RICHLAND, Wash.—The U.S. Department of Energy today launched a new chapter in autonomous biological discovery at Pacific Northwest National Laboratory, commissioning a first-of-its-kind capability designed to accelerate U.S. leadership in the rapidly expanding bioeconomy.

The new Anaerobic Microbial Phenotyping Platform, known as AMP2, marks a major step in redefining how biological science is done in the United States. AMP2 gives researchers the speed and scale needed to accelerate biotechnology discovery and strengthen the U.S. bioeconomy.

AMP2 was commissioned Dec. 4 during a visit to the PNNL campus by DOE Secretary Chris Wright. AMP2 is the newest research capability within the Environmental Molecular Sciences Laboratory, a DOE Office of Science Biological and Environmental Research program user facility located on the PNNL campus. EMSL’s science expertise and sophisticated instruments are used by hundreds of researchers worldwide every year to explore questions about fundamental science.

The official launch of AMP2 marks a milestone allowing DOE and the nation to maintain their global leadership in biotechnology innovation. It’s a burgeoning area: The global bioeconomy is valued at more than $4 trillion but is expected to grow to more than $30 trillion within three decades. The nation’s most advanced autonomous facility for anaerobic microbial research is expected to play a critical role.

Scientists will use AMP2 to explore questions about bacteria and fungi that play important roles in industrial processes to make chemicals, energy, fuels and biomaterials. The capability gives scientists a new window to explore the world of microbes—an invisible yet powerful workforce poised to boost biotech manufacturing as well as provide insights into basic science questions.

Microbial science is a longstanding strength at PNNL, where investigators study their role in producing new chemicals and bioproducts, isolating critical minerals and materials, and managing legacy nuclear waste. At the same time, autonomy and artificial intelligence are reshaping the way science is done, amping up the throughput in laboratories and allowing investigators to accomplish in minutes what formerly took days or weeks.

PNNL is bringing autonomous science, AI and microbial expertise together in a way never done before, positioning EMSL and PNNL as hubs for high-throughput, AI-enabled science. After a competitive award process, PNNL selected Ginkgo Bioworks, a Boston-based company, to build an automated phenotyping platform to the specifications of EMSL scientists. The company, which creates modular, integrated laboratory automation for therapeutics, diagnostics and manufacturing, then built the platform to bring the vision of an automated laboratory focused on microbes to reality.

The AMP2 capability will be housed in an 1,800-square-foot laboratory within EMSL. AMP2 will begin operations by January and will be available to scientists around the nation working with DOE to explore the characteristics of microbes that thrive in environments that lack oxygen. PNNL scientists believe that AMP2 is the world’s largest autonomous-capable science system for anaerobic microbial experimentation. 

AMP2 is a prototype for a larger facility called the Microbial Molecular Phenotyping Capability, or M2PC, which will also be part of EMSL. That facility will include 32,000 square feet filled with more than 100 high-tech analytical instruments for microbial phenotyping, all fully connected and automated. PNNL expects to break ground on M2PC next year, with the facility expected to be ready and open for scientists from around the world to use by 2030. Funding for both AMP2 and M2PC comes from the DOE Office of Science; Ginkgo Bioworks will also build the M2PC platform under a separate competitively bid contract award.

One portion of autonomous science involves robots doing tasks such as filling and loading sample containers and moving containers from one laboratory location to another, operating around the clock. But the systems involve much more than robots doing repetitive tasks. True autonomy also includes designing whole series of investigations so that one experiment flows into another experiment automatically, without constant human intervention.

Autonomy will include the use of AI analytical tools and agents to analyze scientific findings as they happen, then immediately redirect the next experiments based on those findings far faster than any person could. The systems will have the ability to both move automatically to the next round of experiments or to collaborate with scientists to guide experimental designs. At all times, automation engineers will be in the loop, running and maintaining the system.

Automation makes it possible for researchers to do more experiments much more quickly than they otherwise could, giving them a huge dataset from which to draw conclusions far faster and more efficiently than it would take if the work were done by hand. An added benefit is the precision of autonomy; a robot filling a sample container 10,000 times will fill those containers more consistently than human hands can.

PNNL scientist Scott Baker is one of the scientists involved in both AMP2 and M2PC. He’s a fungus aficionado well known in international science circles who likens the projects to major physics experiments where researchers from around the world work together to try to solve basic questions, such as detecting gravity waves or characterizing sub-atomic particles.

“The quest to understand the biology underlying microbial phenotypes is a big deal,” said Baker. “Scientists so far have explored only a small fraction of the microbial species that exist on our planet. To create a very broad, basic understanding of the breadth of the microbial world and to be able to predict and control what these bugs do, we need a huge team of researchers doing impactful science, coupled with one of the most sophisticated laboratories in the world focused on the microbial world.

“With AI and autonomous science, we’re in a perfect storm of technology developments to make disruptive leaps in our ability to advance biotechnology for the bioeconomy,” said Baker. “Our knowledge will go from the equivalent of an old, grainy black-and-white photo of the microbial world to a full-color, 360-degree panoramic image.”

Scientists will look at many characteristics of many microbes—how fast they grow, what nutrients, temperature and pH they like, and how long they produce a compound of interest, to name a few. That collection of traits is known as a microbe’s phenotype, explaining why “phenotyping” stands for the “P” in both AMP2’s and M2PC’s names.

Microbial biotechnology plays an important step in the complex supply chains to make both commodity and specialty chemicals that are widely used in consumer products such as fuels, plastics, cosmetics, food and medicines. Scientists at PNNL and other national laboratories are even studying how to use microbes to extract critical minerals from the ground. 

While bacteria and fungi are used widely in industry today, scientists say much of that work has been the result of trial and error. There are large gaps in our knowledge of how microbes behave under different growth conditions and environments and how scientists might manipulate those conditions to produce more of the products we need more efficiently.

“In the biotechnology sector, biological functions—even in the tiniest of organisms, microbes—are incredibly complex,” said Douglas Mans, interim director of Earth and Biological Sciences at PNNL. “Trying to disentangle the control levers so that microbes do what we want demands a level of experimental technology not currently available. In both AMP2 and M2PC, the level of modular and R&D-focused automation and integration with AI will allow DOE and the scientific community to perform experiments to learn more in ways and on a scale that we simply cannot do using today’s methods.

“This is about much more than simply making current processes more efficient. Automation and AI are vehicles for true scientific innovation. We can perform many more experiments and generate much larger datasets that will lead to new insights that we cannot even imagine,” said Mans, who brought the concept of a large-scale, autonomous laboratory to EMSL when he became its director in 2019.

As scientists modify the growth conditions and environments of the microbes and record what happens, the data will pile up quickly. AI will play an important role in managing all that information, analyzing incoming data instantly and then directing the system to automatically undertake next steps. The results will allow scientists to think more broadly, create new approaches to learn more about microbes, and focus on how to best take advantage of the new information.

“The dream is to gather enough data about microbes under such a variety of conditions that you can order up what you want, and the AI tools will offer a guide to make that happen,” said Todd Edwards, EMSL automation engineer. “That’s the kind of thing we can’t do today, but we’d like to get enough data so that we could do it one day. Microbes will be able to do many things that we haven’t yet discovered.”