Unseen Forces: How Scientists Map the Invisible Interactions of Life

The complexities of life can be broken down into smaller units—ecosystems made up of organisms, organisms made of cells, cells made of interacting molecules, and so on. Scientists have long known that most of the molecules that make up living systems on Earth—like proteins, DNA, RNA, and metabolites—remain uncharacterized. There is also a wide range of interactions among these molecules that remain poorly understood. It’s only in the last few decades, using new methods and technology, that researchers have started to uncover the secrets of this “biological dark matter.”

Armed with some of the world’s most advanced instrumentation, researchers at the Department of Energy’s Pacific Northwest National Laboratory (PNNL) are working to analyze huge amounts of data and uncover hidden biological connections. Not only can they measure biological matter that exists in minute quantities in the environment, but they are also bolstered by powerful computing power. These capabilities underpin a key approach to characterizing biological dark matter, known as network biology, which allows scientists to explore the complex web of molecular interactions in different biological systems.

Exploring biological dark matter

The term “dark matter” often evokes images of the unknown vastness of space, but for Jason McDermott, a systems biologist at PNNL, it has a very different meaning. 

“Biological dark matter refers to molecules whose composition or functions remain a mystery,” he says. “These include proteins, metabolites, and other molecules that we know exist, but whose structures or interactions are poorly understood. These molecules might have profound impacts on human health and the environment, such as disease prevention and crop yields.” 

For example, we might know the structure of a protein but not its function. Or we might be able to identify a metabolite but not know its exact chemical structure. Uncovering these hidden players will help scientists better answer some of our biggest biological questions—this includes everything from evolution and the origins of life to research frontiers like biotechnology, the latter of which aims to develop sustainable bioproducts or custom-made medicines. 

As scientific tools and techniques advance, scientists are now on the cusp of uncovering these hidden networks. This work also has considerable bearing on the emerging bioeconomy which uses biological processes to create new products and services to help society and the environment.

Mapping a different kind of social network

To better understand life’s functions, scientists worldwide use network biology to map biological interactions. They’re trying to answer fundamental questions, such as “Who’s interacting with whom?” and “How do these interactions shape biological systems?” 

To understand how it works, think of human tissue cells or microbes as members of a vast social network. Just as people are interconnected through various relationships, these biological entities are linked through complex interactions. Now imagine this network of complex interactions exists among millions of other networks. 

It’s this network of networks that results in different biological functions and traits, and it’s what scientists are studying to see how different molecules work together—or sometimes against each other—shaping everything from cellular behavior to disease progression. This approach has wide-reaching implications for everything from cancer research to designing new products, such as food and fuel, from living organisms.

PNNL Computational Biologist Sara Gosline, who now finds herself in the midst of a number of network biology projects, didn’t necessarily start her career with an eye towards using this approach in her work. "I’ve always been strong in math and in college I majored in computer science. But I couldn’t quite find my purpose in it,” she says. “When I saw these fields being applied to biology, it was the spark I needed because I saw the tremendous change this work could make for people’s everyday lives."

Gosline’s research focuses on experimental cancer systems. Network biology helps her study how a cancer cell behaves differently from healthy cells with and without drug treatments. This knowledge could eventually lead to drugs that are designed to target those cancer cells by interrupting their networks. These approaches are broadly applicable, including in her work on acute myeloid leukemia with colleagues at Oregon Health & Science University.  McDermott and Gosline both hold joint appointments at OHSU’s School of Medicine, providing them with expanded opportunities for collaboration and a greater impact.

How guilt by association helps uncover molecular interactions

Network biology uses visual and mathematical techniques to map biological data, helping scientists uncover previously hidden functions. This approach can involve a method akin to “guilt by association,” where connections are made between different biological units based on shared patterns or behaviors. This allows researchers to make inferences about how molecules might interact, even before their precise roles are fully understood. 

The beauty of this method lies in its ability to link seemingly unrelated data. “I’m drawn to network biology because we can bring different information into the same context, basically collecting evidence from different sources to answer a single research question,” says Gosline. 

Using these multiple lines of evidence is what allows her to deduce the function of these interactions. “For example, I can measure proteins in one sample and genes in another sample from the same source, such as a cancer tumor. I can then place these in the same graphical network and see if there’s a connection, or a pathway, between them,” explains Gosline. “This connection could help design a drug to interrupt tumor growth and improve patient outcomes.”

In his work on soils, McDermott uses network biology techniques to explore the complexities of microbiomes. Microbiomes are made up of single-cell organisms that rely on each other to survive. They too form networks that can be measured and analyzed. "This research could lay the groundwork for new ways of engineering microbes," he explains. "But first, we need to understand how they fit into the broader network of organisms they’ll coexist with. In essence, we need to understand how microbes interact before we can produce successful bioproducts."

AI as a tool for network biology, not a silver bullet

Network biology is a team effort, and at PNNL, scientists thrive in an environment that encourages cross-disciplinary partnerships. This collaborative approach is becoming increasingly critical particularly because investigating biological dark matter generates vast amounts of complex data.

This is where artificial intelligence (AI) proves valuable, but not as a one-size-fits-all solution. As Gosline points out, AI isn't a "magic wand" for solving all the challenges in network biology. High-quality biological data are not abundant in the portions of the internet where different language models (like ChatGPT) pull from, making it hard to use these tools for biological research. 

"There is still a lot of work needed to organize and filter datasets before they’re ready for AI analysis," she explains. "Once further along, AI has tremendous potential to process and analyze these datasets efficiently." In line with this, McDermott has been working with colleagues at OHSU to advance methods that combine AI with network biology, aiming to unlock new insights from complex datasets.

The combination of AI and biological network representation remains a frontier in biological research. While best practices for integrating AI into network biology are still being refined, different techniques are already showing promise in advancing the field. This approach can reveal insights that weren’t apparent using traditional data analysis methods, especially given AI’s ability to uncover patterns within large datasets.

As network biology evolves, so does our ability to understand the complexities of biological systems. The work being done at PNNL, and beyond, is just the beginning of what could mark a transformative era in biological discovery—one that not only advances scientific understanding but also has the potential to improve the quality of life for people worldwide.