Stephanie Forrest’s unique integration of computation, biology and health

By Luhnyae Campbell

In honor of October being Cybersecurity Awareness Month, we are featuring Stephanie Forrest, director of the Biodesign Institute Center for Biocomputing, Security and Society, and her work combining biology and computation.

At the Biodesign Center for Biocomputing, Security and Society, Stephanie Forrest and her team bridge the realms of biology and computation. The center translates insights between computer science and biology, with a focus on understanding and mitigating malicious behavior in complex systems. For example, what can we learn about cybersecurity from our own immune systems? Can software programs evolve like living things do?  

In addition to leading the center, Forrest studies the biology of computation and computation in biology, including biological modeling of immunological processes and evolutionary diseases, cybersecurity, software engineering and evolutionary computation. She is professor of computer science in the School of Computing and Augmented Intelligence and a fellow of the Institute of Electrical and Electronics Engineers. 

Forrest talks about her unique career journey, the center's pivotal societal impact and the distinctive aspects that set the center apart. Answers are edited for length and clarity.  

Question: What is the research focus of your center? 

Answer: Our center explores deep connections between biology and computation with a focus on understanding and mitigating harmful behavior in complex systems. Systems of particular interest include: biological systems, which are subject to myriad threats such as viruses and cancer; digital systems, which suffer from bugs, malware and misconfigurations; and increasingly, social systems, which have become vulnerable to misinformation, rampant electronic surveillance, locked-in biases and other spillover effects from digitalization. 

Our research leverages these connections by developing novel models of biological processes, creating new bioinspired algorithms, inventing new methods for secure computing tailored to biological data and domains, designing new cancer treatments and protecting freedom on the internet. Because many of the systems we study are distributed spatially or span multiple administrative boundaries, much of our work involves distributed algorithms and the emergence of organization in complex systems. 

Q: Why is this work important to society? 

A: Today we live in a connected world — most of our daily activities, including science and medicine, are mediated by computational processes. Ensuring the safety and veracity of our information and decision systems is crucial, whether it is designing the optimal intervention strategy for my particular cancer, deciding how to vote in the next election or tackling climate change.  

Many of our current technological plagues arose because the people developing and marketing technology were isolated from those who actually use or are affected by it on a daily basis. Examples include machine learning systems that are biased against certain protected groups, big data systems that are not adequately sanitized for errors and omissions, social media platforms that make money by amplifying extremist speech and insecure networks that carry information. Because we are embedded in Biodesign, we have the opportunity to integrate computation with biology and health in novel ways. 

Q: What is the biggest challenge in this field of research? 

A: Fundamentally, we are interested in detecting and mitigating malicious behavior across a wide range of complex systems. We have the capability of bringing analytical theory, modeling and simulation, data-driven empirical work and clinical trials or social experiments to bear on these problems. However, aligning all of these lenses for beneficial predictive intervention is extraordinarily difficult, and communicating the nuance in our work is even more challenging. Deep interdisciplinary work is a contact sport and requires all participants to invest time and effort learning languages and cultures that are specific to other disciplines. This takes time and requires significant upfront investment, which can be frustrating for students and funders. 

Q: What is something you consider one of the center’s biggest successes? 

A: We have improved the operations of three major complex systems: the internet, cancer therapeutics and software maintenance. In the internet space, we uncovered gaping vulnerabilities in the fundamental protocols of the internet that are the foundation of all digital communication. As a result of our discoveries, every major operating system vendor changed how these communications work, and we are all more secure because of it. In cancer, we have succeeded in translating simulation models for a radically new approach to treating cancer to mouse models to early clinical trials. And, in software, we pioneered methods for automatically finding repairs to software bugs, which have been adopted by industry and created a new subfield of software engineering, known as automated program repair. 

Q: How are students involved in the center’s research? 

A: Most of the research in the center is conducted by graduate and undergraduate students. This involves developing research concepts and plans, developing code, running computational experiments and disseminating results. Faculty work directly with students on their projects rather than using the traditional hierarchical large lab model. We recently initiated the BSS Scholars Program, modeled on our existing ACE Scholars Program. This program provides undergraduate students with a research experience in the center, mentored directly by graduate students and postdocs, and a weekly seminar experience with faculty. 

Q: If someone gave your center $100 million, what would you do with it? 

A: I would develop a radically new engineering approach for designing and constructing systems that live in evolving environments. Just a few examples of today's challenges that require radically new approaches include: bacteria that evolve to resist antibiotics, cancer that evolves to resist most treatments, social networks that encourage the spread of disinformation and extremism on the internet, synthetic biology and self-driving cars. These challenges span virtually every field of science and engineering. 

Today's engineered systems are sometimes produced through a combination of human-driven engineering and Darwinian dynamics. In other cases, they interact with evolutionary and adaptive processes once deployed, as in antibiotic resistance, managing invasive species and social media. Yet, we lack a coherent design process to guide the design of such systems. We should ask: “How should engineering practice be changed to account for systems that are designed and deployed in evolving environments?” That is a $500-million question, one that requires radical changes to fundamental engineering principles and education practices. And it is a challenge that could be key to our survival as a species and to saving the planet from imploding. 

The center is already addressing this challenge in a few specific areas, and other isolated examples exist. What is needed is a whole-of-engineering approach, one that generalizes the lessons we have learned and translates them into next-generation engineering practices, which can be taught in the classroom and deployed throughout the various engineering domains. 

Q: Can you share an example of nature-inspired research in your center? 

A: There are several examples, such as evolutionary computation, that are inspired by the mechanisms of biological evolution; comparative oncology, where we identify species with exceptionally low levels of cancer and then discover how they are suppressing the disease; and immune-inspired cybersecurity, which I will explain a bit more here.  

The problem that we face with cybersecurity is similar in many ways to the problem that immune systems solve so elegantly in nature. Immune systems automatically detect and respond to external threats, they use learning to recognize novel pathogens and they use memory to protect against repeated attacks. In addition, they are massively parallel and the cells and molecules that implement the system are distributed throughout our bodies with no central control system. Finally, immune systems are diverse — each individual has genetic and life history differences that convey diverse forms of protection, and even within an individual different immune cells have unique detectors. The architecture of the natural immune system can be abstracted to generate novel cybersecurity solutions, for example, for automated intrusion detection of cyberattacks, for non-cryptographic privacy-preserving data representations, and for automated diversity of host systems (individual computers). This last example is deployed by default on virtually every vendor-supplied operating system. It is known as “address space randomization” and if you are reading this text on a laptop, your computer is most likely using this technique while you read. 

Q: How did you become interested in science, and in particular, the field you are in? 

A: My route was unusual, because my undergraduate institution, St. John's College, is built around reading the Great Books. These included mathematics, from Euclid to Lobachevsky and science, from Ptolemy to Darwin to Einstein. In my senior year, I discovered the beauty and power of Godel's undecidability and incompleteness theorems, which led to the idea of computable numbers and ultimately to computation. Thus, my route to computer science was through mathematical logic rather than a love of tinkering with computers. Using the lens of computation to understand and improve the world around us has motivated all of my subsequent work. 

Q: What key events set you on your research path? 

A: I stumbled into the computer science PhD program at the University of Michigan before computer science was fashionable or lucrative. There, I met John Holland, who became my dissertation advisor. His deep interest and feel for interdisciplinary science resonated with mine, but both of us were quite out of step with the mainline thinking of what was called “core computer science.” His encouragement and vision allowed me to discover my own research voice and stick to my guns intellectually. 

After graduate school, I worked in Silicon Valley long enough to discover that I am passionate about ideas and much less so about the business of making money. From there, I was fortunate to find a home in New Mexico, possibly the only place in the world at that time where it was feasible to think about the nonlinear dynamics of computation (at the Center for Nonlinear Studies, Los Alamos National Laboratory), take biology seriously in a computer science department (at the University of New Mexico) and participate in the creation of a new interdisciplinary science (at the Santa Fe Institute). I didn't have a name for what I have devoted my research career to until I encountered the Biodesign Institute, which describes precisely the vision that my research aims to achieve — uncovering and abstracting the fundamental design principles of biological systems and articulating them in computational systems. 

Q: What is the most fun aspect of your work in the center? 

A: Working with other faculty and students to identify a new research question. The key to successful research is taking the time to formulate the right question to ask — in particular, I like questions that go to the heart of important unsolved problems, questions that will change the world if answered successfully and questions that no one has thought to attempt but that are tractable when looked at in a new way. 

Q: What is your favorite thing about working at Biodesign? 

A: I love the emphasis on collaboration across disciplines, working as a team and the willingness to take risks. This last is important, because it is easy to say “we do risky science,” but quite another to say “and, we tolerate failure when the risks don't pan out.” High-risk research implies that we will often fail, and providing an environment that understands that is key to catalyzing great discoveries. 

Q: Describe your experience with Biodesign’s collaborative, interdisciplinary research culture. 

A: Our center is almost entirely computational, which is unique among the existing centers. In spite of this culture gap, we have more collaboration opportunities than we can handle, everything from working on security and privacy of wastewater data to modeling the microbiome to optimizing molecular dynamics codes. 

Q: Has your teaching and mentorship helped inform your research, and if so, in what ways? 

A: Absolutely yes. My students are usually much smarter than I am. They are expected to act as independent intellectual agents, and my job is to help them refine their vision into an achievable project, encourage them when the going gets tough and point out connections that they may not see. My students have consistently dragged me into new technologies and new subfields of biology, which almost always comes with fascinating new research questions.