As we welcome our 2025 Fellows to Oxford for their first Science Leadership Program convening, we profile Alissa Hummer, who will be returning to familiar ground. 

With a PhD in machine learning from the University of Oxford, the 2025 Fellow is now embarking on a bold interdisciplinary pivot to bioengineering at Stanford University. 

Alissa’s story, and our Science Leadership Program,  exemplify our mission to develop the next generation of interdisciplinary science leaders.

Bridging Disciplines to Transform Human Health

Alissa Hummer clearly recalls the moment she woke up to discover she had been awarded a Schmidt Science Fellowship.

“On that day, it was a holiday in the US, and they must have sent the email while I was still sleeping. And I got up and I ran into the next room to my husband and said, “Can you read this email? Like, am I reading it correctly?” she laughs.

“It really was just a complete game changer. It’s transformed the trajectory of my career, my research, and my life.”

Alissa’s Fellowship Placement is at Stanford University, but this week she returns to Oxford for her cohort’s first Science Leadership Program (SLP) convening.

Hosted in some of the world’s most dynamic and innovative centers of science, the SLP offers bespoke leadership development, including three residential convenings and virtual sessions.

All Fellows participate in the SLP during the first year of their Fellowship.

In Oxford, Alissa will meet the rest of the 2025 cohort in person for the first time, people whom she hopes will become lifelong collaborators and friends.

“I’m excited to be joining this community of amazing, really open-minded scientists and hearing about the science that everyone else is doing. I think it will really broaden my scientific horizons,” she says.

Alissa is European and American, and so far her career has been decidedly British.

After a biochemistry undergraduate degree from Oxford, she completed a master’s at Cambridge’s MRC Laboratory of Molecular Biology in structural bioinformatics and then returned to Oxford for her PhD, where she was supervised by Professor Charlotte Deane MBE (who also serves as a Rhodes Trustee).

For this, Alissa moved deeper into computational research, developing machine learning (AI) methods to aid the discovery of new antibody therapies.

Antibodies are proteins made by our bodies to fight pathogens and disease.

Since the 1980s, it has been possible to make them in the lab and deliver them to patients as therapeutics.

They are now the best-in-class therapies for many diseases, from cancers to Covid. But each takes a decade or more to develop, at a cost of millions of dollars.

Alissa’s doctorate focused on better prediction of the strength with which antibodies bind to antigens, their target proteins found on the surface of cells and viruses.

This is known as binding affinity. Its complexity makes predicting binding affinity one of the biggest bottlenecks in the therapeutic antibody development pipeline.

She developed an AI model, Graphinity, that performed well at predicting antibody-antigen interactions that were similar to data that it had been trained on, but performed poorly when it came to predicting entirely new interactions (so-called generalizable prediction).

To discover if this was due to the model itself or the training data, Alissa used synthetic data – computational simulations of real experimental data – to train Graphinity on over a million possible variations of antibody-antigen interactions.

Her findings, published in Nature Computational Science, revealed a stark fact: generalisable prediction by AI would need hundreds or thousands times more data than we currently have available.

Importantly, her research reconfirmed the promise of AI for accelerating antibody therapy development when enough good-quality data is eventually available.

It also highlights other urgent priorities, such as the need for more varied, well-structured, and standardised datasets to train machine learning models.

But Alissa is also curious about the bigger picture, dreaming of one day being able to build 3D models of cells as a tool to study the causes of diseases.

Now, as a Schmidt Science Fellow, she is embarking on a postdoctoral placement at Stanford University, where she has pivoted to wet lab science to model cells and their biology in a way that better represents real life.

“In my PhD, I was modelling molecules completely in isolation,” she explains. “But in real life, in nature, that’s not how molecules look. They’re surrounded by a very complicated environment. And there are lots of different factors that right now we aren’t considering at all in our computational models.”

She will use bioengineering and microscope imaging techniques to explore how stem cells differentiate into many other types of cells, in both real human cells and virtual ones.

Like many researchers pursuing an interdisciplinary path, there are challenges to learning a new field.

Alissa hadn’t set foot in a wet lab in seven years and had needed to learn new experimental procedures and tools.

Her pivot science requires her to think on a different scale, gathering data for entire cells rather than individual molecules.

Her Schmidt Science Fellowship provides a unique opportunity to do all of this.

“The Schmidt Science Fellowship will actually make my research stronger in the end. It’s not just a deviation in my path or a curiosity for a couple of years. This will make me a better scientist and able to have more impact.”

“This is a big pivot from my PhD,” Alissa explains.

“The Schmidt Science Fellowship will actually make my research stronger in the end. It’s not just a deviation in my path or a curiosity for a couple of years. This will make me a better scientist and able to have more impact.”

Alissa’s interdisciplinary work, developed in world-leading environments on both sides of the Atlantic, could one day lead to next-generation treatments that improve many different aspects of human health.

“The Schmidt Science Fellows program, delivered in partnership with the Rhodes Trust, fosters adventure and ambition in the most talented early career researchers, and it plays a vital role in enabling them to become the interdisciplinary science leaders of tomorrow,” Professor Deane says.

“Having seen Alissa’s growth during her PhD at Oxford, it is inspiring to see her pursuing her research through this bold interdisciplinary pivot.  This new direction will open rich avenues of inquiry, and it’s exciting to imagine where Alissa will take it.”

Alissa is now taking the next steps in her research journey and is looking forward to forging connections in her new Schmidt Science Fellows community that will challenge her thinking, accelerate her discoveries, and build friendships that last a lifetime.