1 min readResearch Matters

‘The current strategy for dealing with drug resistance is like Whac-A-Mole’

Chemical engineer Brian Hie is using machine learning to predict how diseases will evolve and develop drugs that can treat them even as they change.

Headshot portrait of Brian Hie
Brian Hie at laptop, looking at image of genetic code
Closeup of image on screen, with targeted area highlighted

In the “Research Matters” series, we visit labs across campus to hear directly from Stanford scientists about what they’re working on, how it could advance human health and well-being, and why universities are critical players in the nation’s innovation ecosystem. The following are the researchers’ own words, edited and condensed for clarity.


Diseases are a moving target because biology is constantly changing. Tumors evolve resistance to cancer drugs. Drugs made of protein antibodies are used to combat viral diseases, but viruses change rapidly, so if a virus develops a resistance to the protein therapeutic, then the drug is completely useless. A billion dollars will be spent to develop another drug, which is equally likely to encounter future resistance.

In other words, the current strategy for dealing with drug resistance is like Whac-A-Mole; a new resistance mutation pops up, we try to address it, but ultimately, we’re taking a purely reactive response to evolution.

My research group is focused on two areas. First, we try to use machine learning to better predict and simulate evolution of diseases. If we can predict at what point a disease will mutate to evolve resistance to a drug, we can potentially design protein therapeutics that will work against the future form of that disease and get a head start on dealing with viral evolution. Our goal is to get really good at predicting pathogen or disease evolution so we can develop drugs that resist diseases as or before they evolve.

Second, we are focused on developing better models of biology that will allow us to better design and control biological systems. There’s no point in understanding disease evolution if we don’t have the tools to cure or prevent the disease. Currently, our goal is to expand the modeling capabilities beyond individual molecules to the level of entire biological systems. This is important because most diseases are caused by complex systems beyond the molecular level. Our newest models can now understand biology at the level of entire organisms.

When I would watch movies as a child, the people I admired were the ones with all the cool gadgets and technologies. But I didn’t want to actually be James Bond, I wanted to be like Q or whoever was giving James Bond all the technology. That is really what drew me to research. It seems like magic, in that it takes us beyond our current capabilities, but also not like magic, in the sense that it’s reproducible and you can achieve it again and again.

This work is really hard to do and it requires a lot of training. If you want to get really good at AI or biology or any other form of scientific research, there are very few places outside of academia that allow you to focus just on that. We need institutions that train and teach people these very complex skills.

It requires courage to accept data that we don’t think aligns with our beliefs, but that’s the whole point of science – it challenges our beliefs.”

It’s worth the investment. Many people can benefit from the fruit of biomedical research. The economic value created by new companies that have come out of biomedical research is orders of magnitude more than what it costs to train those scientists.

Some people might think of science as being tied to one or another political agenda, but science just follows the data. It requires courage to accept data that we don’t think aligns with our beliefs, but that’s the whole point of science – it challenges our beliefs. It’s a big misconception that scientists have all the answers. It would be great for more scientists to communicate that a lot of science is uncertain. It’s a process that’s always rewriting itself. It’s a very dynamic and active thing.

For more information

Brian Hie is an assistant professor of chemical engineering, a member of Stanford Bio-X, an affiliate of the Stanford Institute for Human-Centered Artificial Intelligence (HAI), a faculty fellow of Sarafan ChEM-H, and the Dieter Schwarz Foundation SDS Faculty Fellow at Stanford Data Science.

Photographer

Andrew Brodhead

Share this story

Research Matters

Groundbreaking innovations that begin in Stanford labs flow freely into private industry to improve human well-being, fuel the economy, and strengthen American competitiveness.
Learn more