�԰�����

On a weekly basis, Cole McGuire ‘25 stands at an intersection of two different worlds.

����computer science major who came to �԰����� with a passion for��biology, he’s using this unusual combination of interests to conduct interdisciplinary research that could have huge implications for how machines (namely, computer algorithms) can solve some of humanity’s biggest biological mysteries.

“I have a passion for research in two things that seem in some ways almost polar opposites: the natural world and the artificial one. Trying to get those to interact in a way that can be beneficial to us has been just fascinating,” McGuire says.

At �԰�����, McGuire works alongside an interdisciplinary duo of computer science professor��Matthew Hibbs, Ph.D., and biology professor��Bethany Strunk, Ph.D. Now in a third year of research, the team’s project, “A Deep-Learning Approach to Gene Function Prediction,” is building on a tantalizing way to use AI to speed up the human understanding of genes, which in turn has potentially crucial applications for fields such as cancer research.

Here’s the project abstract: Genes are segments of DNA that command various parts of your body to perform various duties. Humans have about 20,000 genes in total, and we still don’t know what they all do. Testing each one individually is impossible, and working through the sheer number of potential combos to figure out just what goes right and wrong (such as with cancer) is a labor-intensive, expensive, and, frankly, superhuman task for biologists.

Cole McGuire '25 (left), alongside biology professor Bethany Strunk, Ph.D. and computer science professor Matthew Hibbs, Ph.D.

The time scale of doing this manually would be “not just one or 10, but 100 lifetimes,” Hibbs remarks. “But with artificial intelligence,” Hibbs says, “we can find correlations between the levels of genes in our bodies, and we can use that data to predict what the genes are actually doing. When we can predict what a gene does, we can study specific diseases without having to go through the whole testing process one by one of all 20,000. And maybe we can narrow down the search space to something that’s more manageable to do with our experimental colleagues.”

And thanks to Strunk, the team just needs to start at 6,000, which is the number of genes in yeast, a workhorse research organism that allows this type of research to happen more efficiently and without needing to use human cells.

“Our ability to manipulate yeast genetically is so simple and so fast,” says Strunk, who hosts McGuire in her lab as he switches back and forth between computer science and biology modes weekly. “This is an awesome learning experience and way to interact with people that I might not otherwise get the opportunity to interact with and to see if they can help biologists answer the questions that we’re asking.”

There’s a name for this crossover field: bioinformatics. “I like to think of it as trying to solve biological problems using the tools and abilities of computer science,” Hibbs explains. “And that’s become more important in the genomic (gene-studying) era because the amount of data and the number of questions that biology has right now is greater than any one person can hold in their head.”

Cole McGuire '25 brings a unique blend of computer science expertise to his research, which has huge implications for the field of biology, another passion of his.

McGuire was seemingly destined to excel in this bioinformatics field as a student at �԰�����, where the��liberal arts put students seeking hands-on opportunities that bridge disciplines in a position to thrive. Alongside his computer science major, he’s a biology minor (just a few classes short of a major) and a��philosophy minor to boot.

“With computers, we’re able to achieve and do some things that feel like magic,” McGuire says of his passion for computer science.

“Whereas with biology, it is fascinating to take a glimpse into...these extremely complex systems that are all working in harmony to produce what just seems average to us. Maybe we’ll never come to a full understanding of the human body, but just uncovering a little bit of these threads that wind in specific places can improve our quality of life so much.”

McGuire decided to enroll at �԰����� thanks in part to a tour that gave him a peek into a computer science course with professor��Mark Lewis where students were excited to participate in class. And upon arriving on campus, McGuire took a��First-Year Experience (an interdisciplinary class for new Tigers) in science fiction where he met many of his future classmates in biology, computer science, and other��STEM fields.

Just three weeks into his college career, McGuire found himself reaching out to Hibbs to see whether there was a way he could start doing research that bridged biology and computer science.

“I was a bit nervous. I’m just a first-year who’s taken one computer science class. I haven’t taken a biology class,” McGuire says with a laugh. “But he recommended a few papers that I could read to get an understanding of the field, and then just a few weeks after that, he asked me if I would like to do some research.”

Cole McGuire '25 and computer science professor Matthew Hibbs, Ph.D. confer about their research into how AI can help speed up biological discovery.

Finding someone like McGuire was a godsend for Hibbs, who needed a student with an interdisciplinary perspective; In short, he needed a �԰����� Tiger.

“When you get that student who understands both worlds, there are different problems that they can tackle,” Hibbs says of McGuire’s dual interests. “If I was working with just a computer science student, we might be thinking about a really technical aspect of machine learning or optimizing some algorithm. But with a student like Cole, we’re really able to analyze the results of what we’re predicting, to think about the biological quality of those predictions. You get to ask a whole different set of questions with someone with that interdisciplinary background and desire.”

Working with both Hibbs and Strunk has evolved past the point of nervous emails for McGuire.

“We’ve gotten to where Dr. Hibbs feels very much like a colleague,” McGuire says. “He is very collaborative, extremely open to experimentation and trying things that are just super off the wall. It might be a bit intimidating sometimes for a student to go to their professor and say ‘I have this crazy idea of this new thing that we can do,’ and Dr. Hibbs is just like, ‘Let’s try it.’ And that’s always been a super exciting prospect.”

And partnering with Strunk on the biology side of the project has been a plus for McGuire, who says he feels honored that he gets to work in a bio lab without having to be a biology major.

“Dr. Strunk is one of the friendliest and most kind people I’ve ever met,” he says. “The way this really applies to research is that she ... is someone who is able to take tons and tons of ideas and connect them in ways that are also off the wall, but she can keep them all on her mind. This allows for such a rich discussion that you’ll have with her.”

Biology professor Bethany Strunk, Ph.D (L) and Cole McGuire '25 examine a yeast sample as part of their research.

As they cross these disciplinary lines, this research team is doing work that even five years ago would seem truly futuristic. But Hibbs says the idea of using artificial intelligence (AI) to cut down on gene research time is not new—people just have better tools now.

“Neural networks, which is what we call ‘deep learning,’ are a very old idea that dates back to the 70s, but they sort of went out of fashion because they’re very computationally intensive to train and run, and they require a lot of training data,” Hibbs says. “The good thing with genomics is we have a lot of data, and now with a lot of the (technological) advancement... our idea was to return to these older methods and see if these newer technologies and developments in neural networks could outperform what we were doing back in the day.”

Both McGuire and Hibbs are guardedly excited about AI’s potential for their field.

“AI has not gotten to the point where it really scares me in terms of its potential, but I do think you’re starting to see it dipping its toes into every area (of our lives),” McGuire says. “Being someone who’s working on a very specialized use of AI in a non-computer-science field shows me the wide array of potential it has, but it also shows me some of the limitations, and it shows me the places we still have to go.”

And as helpful as AI has been to the team’s work, it will never replace the most powerful tools the group has: a natural sense of curiosity, collaboration, and passion for the journey of research, not just the end result.

“When you’re in your classes, you’re learning ‘the known’—what our knowledge of the field is currently. But what makes research exciting is the unknown. There’s not a set answer in front of you,” McGuire says.

That fascination, to professors like Strunk, is just as much of a valuable result of undergraduate research at �԰����� as any thesis or headlining publication.

“Sharing a student’s excitement, that type of teaching becomes a reward in itself, even more than the productivity of making some kind of splash in the field or maintaining some kind of reputation,” Strunk says. “It really makes my life a lot happier that this is a big part of the job.”

��

��

��