Mary Cundiff is a postdoctoral researcher at the University of Pittsburgh. She uses machine learning and single-cell genomics to discover shared biological mechanisms across diverse organs to improve multi-organ therapeutic strategies. In this Postdoc Portrait interview, she shares the nuances that inform her work at the organ, tissue, and genetic levels.
Decoding Multi-Organ Pathology with Machine Learning
Q | What drew you to machine learning in biology?
I was initially drawn to research by the realization that there are still fundamental questions in biology that we cannot yet answer. That sense of open-ended discovery, standing at the edge of what is known, has stayed with me throughout my career.
My path into this specific field came from a combination of neuroscience, immunology, and data science. I became increasingly interested in how complex biological systems break down in disease, and how we can use computational tools to make sense of that complexity. What especially motivated me was the gap between the amount of data we can generate and our ability to interpret it in a meaningful way.
This led me to focus on interpretable machine learning in biology. Rather than treating models as black boxes, I wanted to build approaches that reveal underlying biological structure. Studying disease across tissues felt like a natural extension of that goal because it forces us to ask not just what changes, but what patterns are truly fundamental.
Q | What scientific problem are you trying to solve?
I study how diseases like fibrosis develop across different organs, and whether the same underlying biological programs drive disease in the heart, lung, kidney, and beyond. While many studies focus on a single tissue, patients often experience systemic disease, and we still lack a clear understanding of what is shared versus tissue specific.
To address this, I develop interpretable machine learning approaches that combine single-cell genomics with modern AI models. My goal is not just to identify which genes change in disease, but to understand whether those genes participate in the same biological programs across tissues. By quantifying this “sharedness,” we can begin to map conserved disease mechanisms rather than isolated observations.
Ultimately, I aim to move beyond descriptive biology toward a more unified view of disease, one that reveals common pathways that could be targeted across multiple organs. This has the potential to improve how we study complex diseases and identify therapeutic strategies that are broadly effective rather than tissue specific.
Systemic Biology and the Future of Target Discovery
Q | What’s one thing you learned from your systemic approach that you didn’t expect?
One unexpected lesson was how difficult it is to define what it means for biology to be “shared” across systems. Intuitively, it seems straightforward. If the same genes change in multiple tissues, they must be related. But in practice, the situation is much more nuanced.
Genes rarely act in isolation; they function within coordinated programs. I found that two tissues can show similar gene-level changes but organize those genes into very different biological structures. Conversely, tissues can use different genes to accomplish similar functional outcomes.
This shifted my perspective from focusing on individual genes to focusing on patterns and relationships between them. It also reinforced the importance of building models that capture structure, not just signal.
More broadly, it taught me that biological similarity is not just about overlapping; it’s about how systems are organized. That realization has shaped how I design both my computational methods and the biological questions I ask.
Q | If your research succeeds, what could it change for science or society?
This work could change how we study complex diseases by shifting the focus from single-tissue analysis to shared biological mechanisms across the body. Instead of developing treatments that target one organ at a time, we could identify conserved disease programs that are relevant across multiple tissues. This could lead to more efficient drug discovery, better prioritization of therapeutic targets, and a deeper understanding of systemic diseases like fibrosis, cardiovascular disease, and chronic inflammation. More broadly, it could help bridge the gap between large-scale data generation and actionable biological insight, making modern genomics more useful for both basic science and clinical translation.
Q | What question are you most excited to answer next?
I am most excited to understand whether conserved disease programs can predict treatment response across tissues. If a biological program is shared between the heart and lung, for example, can targeting that program improve outcomes in both organs? This question moves beyond identifying patterns to testing whether those patterns are functionally meaningful.
Responses have been edited for length and clarity.
Are you a researcher who would like to be featured in the “Postdoc Portraits” series? Send in your application here.
