The Humans of the Wyss (HOW) series features members of the Wyss community discussing their work, the influences that shape them as professionals, and their collaborations at the Wyss Institute and beyond.
In many ways, comparing mice to humans is like comparing apples to oranges. Yet animal models have been a huge part of the preclinical pipeline for therapeutics development for nearly a century. With limited quantifiable ways to compare species, it’s no wonder that a vast majority of medicines fail in clinical trials. So, Harvard Bioengineering Ph.D. candidate and Wyss computational researcher Joshua Price decided to try to nail down the differences between species and build models to predict the accuracy of preclinical data, eventually lowering clinical trial failure rates and therapeutic development costs. Now, he’s leading the IMPACT Validation Project to further de-risk, refine, and validate his methodology. Learn more about Joshua and his work in this month’s Humans of the Wyss.
What are you working on?

I’m leading a Validation Project called IMPACT, which stands for Interspecies Modeling for Predictive AI-driven Clinical Translation. For this project, I am developing a computational methodology to forecast whether new drug treatments that work in mice will also work in humans. Specifically, I’m training machine learning models, which are a specific kind of artificial intelligence (AI), to take single-cell RNA sequencing data from preclinical studies in mice and use generative AI to predict how the treatment may look in humans, along with confidence measures for that prediction. The project involves computational and wet lab work, since I am both using publicly available data and generating my own. This system would complement other new approach methodologies, also known as NAMs, being developed and used at the Wyss, such as Organ Chips and organoids.
What real-world problem does this solve?
Animal models, typically mice, are a huge part of the preclinical pipeline for drug development, but they have limited predictive power. That’s one of the reasons that the clinical success rate for new drugs is pretty abysmal: only about 10% of drugs that start phase 1 clinical trials pass their phase 3 trial. To reach phase 1, they need promising preclinical data. So, IMPACT’s big-picture goal is to give researchers tools to better choose which drugs are good candidates for clinical trials and improve their success rate. By doing so, we also hope to lower drug costs. Right now, for every one drug that succeeds, a company has to pay for nine drugs that fail, so to recoup that cost, the effective medication ends up being exponentially more expensive for patients and healthcare payers.
Right now, the first applications I’m working on are vaccines, specifically tuberculosis vaccines, and checkpoint blockade inhibitor therapy for head and neck cancers.
What inspired you to get into this field?
When I first started as an undergraduate at the University of California, Berkeley, I studied mechanical engineering. I really enjoyed the technical problems, math, and systems-level thinking. I liked quantifying and solving problems with ingenuity. But as I thought about my contributions to the bigger picture and the world around me, I was drawn to medicine. It’s an area that will always be important. Regardless of what happens economically or politically, as long as there are humans, we will care about preventing and curing diseases and treating people who are ill.
Even though I’m not working with patients firsthand and don’t want to become a medical doctor, being just one step removed from patients helps me stay connected to the problems we’re trying to solve and the people we’re ultimately doing the research for
I ended up adding a second major in molecular and cellular biology but approached it from an engineering lens. I wanted to use our knowledge of Nature and biology to inspire solutions to real problems. Then I got my master’s in bioinformatics and worked as a software engineer before beginning my Ph.D. in bioengineering.
What continues to motivate you?

It’s very motivating to be in the Wyss ecosystem, where we sit at the border of Harvard Medical School, both academically and geographically. That means there are doctors all around. For example, I sit next to Yoav Binenbaum, a pediatric oncologist at Boston Children’s Hospital. We talk every day and work closely together. Even though I’m not working with patients firsthand and don’t want to become a medical doctor, being just one step removed from patients helps me stay connected to the problems we’re trying to solve and the people we’re ultimately doing the research for. It reinforces that the work I’m doing is important and could make a meaningful difference for people, like Yoav’s pediatric cancer patients.
What excites you the most about your work?
The most exciting part of this project, beyond the potential impact, of course, is that, intellectually, I think the approach we’re taking is new and interesting. It’s focused on quantifying interspecies biology, or the differences between species. Right now, we know a system is similar in mice and humans, but beyond saying whether they have similar genes, we’re really limited in making quantitative comparisons. This project is entirely quantitative on a phenotypic (rather than purely genetic) level, and we have early signs of success. I can see it leading to valuable discoveries that help us understand which therapeutic approaches translate between mice and humans and which don’t. This isn’t restricted to one specific drug; it’s a fundamental idea that could potentially generalize to a variety of therapeutic areas.
What are some of the challenges that you face?

The biggest challenge is wrangling all of the data. I use a ton of data to train my machine learning models. For the most part, it’s publicly available single-cell RNA sequencing atlas data with hundreds of thousands to millions of cells, each containing a ton of information. Single-cell RNA sequencing data is complex: you can apply many transformations, and there are many ways to label what you’re looking at. I need to keep track of all of that across studies, cell types, and species. IMPACT is too complex to write scripts to do data analysis; it requires a full Python software package.
Why did you want to pursue your Ph.D. at the Wyss?
The Wyss is known as both a great place to do research and a great place to see your discoveries or inventions translated beyond academia. The Institute is a big contributor to Harvard’s patent portfolio. While publications are important, patents are even more essential for getting your technology out into the world. That’s my ultimate goal: to develop something that industry and patient care adopt.
While publications are important, patents are even more essential for getting your technology out into the world. That’s my ultimate goal: to develop something that industry and patient care adopt.
What is unique about the Wyss? How has that impacted your work?

The Institute is truly interdisciplinary, and not just as a buzzword. I’m part of David Mooney’s lab, which is already multidisciplinary. I’m a computational person, but we have doctors, materials scientists, and protein and tissue engineering specialists. If I need support in any research area not covered by our lab, there is someone at the Wyss who specializes in anything I have questions about. Here, it feels like we have all the different departments you would traditionally see separated at a university spread across just three interconnected floors in one building. That’s really cool and makes my work easier and more efficient.
How do you collaborate with and/or receive support from teams across the Wyss Institute?
On the Validation Project team, we have people that I would have worked with if it were traditionally funded, for example through an R01 grant, plus additional people at the Wyss who are serving different functions. Senior Scientist Megan Sperry is a computational researcher who co-leads the Translational AI Catalyst. She’s been connecting me with other computational researchers, both at the Institute and beyond. Ally Chang is the business development representative for our project. She’s connected me with contacts at pharmaceutical companies to discuss how the type of technology I’m developing could be used and what their real needs are. By pushing me to work with people outside of just my lab, the Validation Project program has broadened my vision for IMPACT.
How have your previous work and personal experiences shaped your approach to your work today?
My experience as a software engineer gives me a different perspective on reproducibility, repeatability, and robustness when it comes to the complex biological data we capture in research. We’re not just capturing a single number; we’re capturing a million numbers at once, and tons of variables dictate what those numbers mean. There are multiple ways to interpret the same data depending on what tools or transformations you use. Plus, when I’m using publicly available data from different groups and studies, it can be difficult to compare them. You need to think systematically about what the data is, how it was generated, what differences may exist, and then how to compensate for those differences. As a software engineer, I am used to thinking about how to process messy data in a way that is reliable and repeatable. This enables me to test whether my assumptions are quantitatively sound and think about how to make my predictions more robust and accurate.
What do you like to do outside of work?

1/5 Joshua enjoys exploring New England with friends. Here, he grabs lunch in Portland, Maine with fellow Mooney lab members Tania To, Emily Roloson, and Olwyn Mahon. Credit: Joshua Price 
2/5 Joshua is learning to sail on the Charles River. Credit: Joshua Price 
3/5 Joshua and fellow Mooney Lab members Wei-Hung Jung, Tetsu Harimoto, and Kyle Ruark take a pit stop on a bike ride up Cape Cod. Credit: Joshua Price 
4/5 Joshua had the chance to stand on a lake in Vermont on a weekend trip. Credit: Joshua Price 
5/5 Joshua takes a walk with his cat Jiji on the Boston Harbor waterfront. Credit: Joshua Price
I am from the San Francisco Bay Area in California, so although I’ve been here for four years, I’m not a native New Englander. I like to explore the area with friends. There is a ton to do in Boston, whether it’s visiting different neighborhoods, parks, or museums. Outside of the city, there are many towns, beaches, and mountains. Recently, the foliage has made everything really beautiful. I enjoy taking a day trip, or even a half-day trip somewhere new with my friends to experience this part of the country.
What’s something unique or fun about you that someone wouldn’t know from your resume?

I was a Joshua tree for Halloween this year. Only about half of the people who saw my costume knew what a Joshua tree was, because they don’t exist on the East Coast, only in a tiny part of Southern California. I made the costume from scratch, gluing 550 golden tassels and making a leafy crown and props. It took about six hours to assemble.
If you had to choose an entirely different career path, what would it be?
It’s not totally different, since I might be a professor someday, but I could see myself becoming a teacher. I enjoy learning and sharing what I know about the world. In a way, it’s a different kind of discovery. As a Ph.D. student, I’ve had the chance to teach as part of my training. I was a teaching fellow in Dave Mooney’s tissue engineering class, which was fun. I taught students how to do tissue culture and think about their own little experiments. I was also a teaching fellow for a data science class, which was a totally different but still enjoyable experience. I led a section where students were coding and did computational projects.
What does it feel like to be working on cutting-edge technology that has the potential to have a real and significant impact on people’s lives and society?
Knowing that I’m working on these deeper and timeless health-related problems for humanity, while also having the chance to think about fundamental biology, is really fun and meaningful.
It’s exciting and motivating to come to work. Knowing that I’m working on these deeper and timeless health-related problems for humanity, while also having the chance to think about fundamental biology, is really fun and meaningful. It means I am thinking about life, how humans work, and how animals work. It enlightens me about the world in which I live. It’s wonderful to use my knowledge to help people.
