Scholar in Residence aims to further synthetic biology engineering research
Above: An automated robotic pipetting system can process hundreds of samples quickly for experiments that include DNA assemblies and PCR reaction setup with increased accuracy and precision.
Myers at the Living Measurement Systems Foundry, an automated facility for testing and measurement of engineered microbes, located at NIST headquarters in Maryland.
When engineers design a bridge or a computer chip, they rely on proven design principles.
Biology doesn’t always play by those same rules given its complexity.
Professor Chris Myers is partnering with the National Institute of Standards and Technology (NIST), with support provided by the National Science Foundation and Schmidt Science, to develop data standards that could make synthetic biology more predictable and help artificial intelligence become a more effective research tool.
This is one of the challenges driving Myers, professor of electrical, computer and energy engineering and Palmer Leadership Chair at ĢӰ Boulder. During his yearlong sabbatical, Myers is collaborating with NIST researchers in Gaithersburg, Maryland to improve how scientists organize and analyze biological data, laying the foundation for more reliable synthetic biological engineering.
His work combines decades of research in synthetic biology with new opportunities in machine learning and artificial intelligence. Scientists need better data before artificial intelligence can accelerate scientific discovery.
“Engineering works because we can trust our models,” he said. “The more we can bring that same predictability to biology, the more powerful the field becomes.”
We spoke with Myers about synthetic biology, engineering standards and why AI is only as powerful as the information it’s given.
What is synthetic biology for those who may not be familiar with the field?
Synthetic biology involves redesigning organisms for useful purposes by engineering them to have new abilities.
Scientists are harnessing synthetic biology to solve pressing problems in medicine, environmental remediation, manufacturing and agriculture.
The goal of synthetic biology is to bring engineering discipline into biological design. Engineers rely on three fundamental ideas: standards, abstraction and decoupling.
Standards allow people to communicate and build systems that work together. Abstraction lets us design complex systems without worrying about every individual component. Decoupling separates design from construction.
Synthetic biology aims to bring those same engineering principles to biology. The challenge is that biology is incredibly complex, so we’re still working toward making those engineering concepts more practical and reliable.
Why are standards so important in synthetic biology?
Without standards, every laboratory ends up doing things differently. You can still conduct excellent research, but it’s difficult to reproduce results, share data or build on someone else’s work. Standards provide a common language that allows researchers to exchange information in ways both humans and computers can understand.
I’ve spent over 15 years helping develop the Synthetic Biology Open Language, which is one standard designed to improve how biological designs and experiments are represented digitally.
Developing standards is only part of the challenge, though. The harder part is getting the broader research community to adopt them.
How does your research endeavors at NIST help advance the field?
Myers demonstrating pipe petting at a synthetic biology outreach event for community college, undergraduate and graduate students.
NIST has built an impressive automated laboratory for synthetic biology research. Instead of scientists manually performing every experiment, robotic systems like the Living Measurement Systems Foundry can prepare samples, move them between instruments and collect data over many hours with minimal human intervention.
My project has three primary goals. First, we’re integrating the data infrastructure my research group has developed into NIST’s experimental workflows. Rather than organizing information after an experiment is complete, we want every step recorded in standardized formats as experiments happen.
Second, we’re asking a fundamental question: what data should scientists be collecting? That’s actually one of the biggest challenges in synthetic biology. We can often make something work in the lab, but when we move it into a real-world environment, its behavior changes. We want to establish better ways to characterize biological parts so researchers can predict how they’ll perform under different conditions, like how engineers understand the performance of electronic components.
Finally, we’re exploring how machine learning can help improve predictive biological models using these standardized datasets.
Artificial intelligence is becoming part of nearly every scientific discipline. How do you see it fitting into the world of synthetic biology?
Machine learning has existed for decades. What’s changed is that advances in computing and large language models have made it much more accessible. The important thing to remember is that AI doesn't create understanding by itself. It analyzes large amounts of data and identifies patterns.
If your data isn’t well organized or if your model simply memorizes the data instead of learning meaningful relationships with the data, you can end up with predictions that don’t hold up outside the original experiment.
We want machine learning to help us improve our scientific models by showing us what might be missing. Ideally, it points researchers toward new biological hypotheses that can then be tested experimentally. Ultimately, scientists still need to understand why a model works.
You often compare biology to traditional engineering disciplines. Why is predictability so important?
Engineering depends on trust. When civil engineers design bridges, they trust that the models they’re using accurately predict how the bridge will perform.
Biology hasn’t quite reached that point yet. We still don’t fully understand how genetic systems behave when they’re placed in different organisms, different environments or different conditions.
The more we understand those behaviors and the more consistently we measure them the better we'll be able to design biological systems that perform as expected. That’s where engineering principles can really transform biology.
How have your previous sabbaticals informed your research endeavors?
They’ve been transformative. My first sabbatical introduced me to synthetic biology and completely changed the direction of my research career. Another sabbatical led directly to collaborations that became the data infrastructure we’re using today.
This current one is giving me the opportunity to learn how machine learning can genuinely accelerate scientific research because we want to understand where it adds real value. Sabbaticals allow faculty to step away from their daily routines, learn something new and bring that knowledge back to their students, research groups and the university.
If I finish this year having expanded my own understanding and developed new tools that other researchers can use, then I’ll consider this a major success.