By Emily White
Dr. Yang Lu, Assistant Professor in Biomedical Engineering, joined UW-Madison recently as part of the RISE AI initiative, aimed at seeding collaborative opportunities in Artificial Intelligence and data science. Dr. Lu was recruited from the University of Waterloo in Canada, where he was an Assistant Professor for two and a half years. His research is developing interpretable AI methods to make biomedical and genomic discovery more efficient for biologists.
Lu attained his undergraduate and Master’s degrees at Shanghai Jiao Tong University. Pulled by the allure of software engineering, he interned at Microsoft before pivoting to computational biology. Having realized that biology was data-rich but not discovery-efficient, Lu studied computational biology at the University of Southern California, working with Dr. Fengzhu Sun on “big data” analytics in metagenomics. He went on to conduct postdoctoral training with Bill Nobel in Genome Sciences at the University of Washington in Seattle. His work has focused on everything from modeling microbiome traits with taxonomy-adaptive neural networks to multi-modal single-cell data integration to interpreting peptide sequences in mass spectrometry data.
Lu is particularly interested in the collaborative environment at UW-Madison. “Some of the most useful research ideas come from seeing connections across different fields. Since my work sits at the intersection of AI, statistics, and biology, staying curious beyond my own discipline has been surprisingly valuable,” he says. We asked him about his research program and trajectory.
Describe your research interests and program.
My research program sits at the intersection of AI, statistics, and biology. Broadly, I am interested in how we can use artificial intelligence to help biomedical researchers move from large, messy molecular datasets to clear and testable biological hypotheses. Modern biology is incredibly data-rich, but that does not automatically make discovery easier. In many cases, the real challenge is figuring out which signals in the data are meaningful and worth following up experimentally. My lab develops methods that help scientists find those high-value signals more efficiently and with greater confidence.
The main goal of my research is to make biomedical discovery more efficient, more reliable, and more interpretable. Instead of relying only on slow trial-and-error hypothesis generation, I want to build AI systems that can learn from large-scale data, prioritize the most promising genes, pathways, or molecular patterns, and tell us not just what might matter, but how confident we should be in those suggestions.
What system or technology do you focus on?
My lab focuses especially on multi-omics data, including high-dimensional measurements collected at the level of individual cells. We also work on methods for integrating datasets collected across different labs or technologies, so that researchers can learn from larger and more diverse data collections without being misled by technical differences. A third major area is interpretable and scientifically useful AI: building models that do not function as black boxes, but instead produce biologically meaningful explanations and a short list of experimentally actionable leads. I want to help shift biomedical research toward a more data-driven and AI-guided paradigm.
How does your work relate to the RISE AI initiative?
RISE AI is a particularly strong fit for my research because my work depends on close interaction between AI development and real biomedical applications. The initiative gives me access to an interdisciplinary community across engineering, medicine, and biology, which is exactly the environment needed to turn AI methods into useful scientific tools. It also strengthens the translational side of my work by creating opportunities to connect methodological innovation with real disease questions, experimental collaborators, and longer-term clinical impact
Is there a single person or experience that most influenced your trajectory to where you are today?
Yes—my postdoctoral advisor, Bill Noble, has been one of the most important influences on my trajectory. He has been a role model for me not only because of the depth and impact of his research, but also because of how he leads a lab and mentors people. What I especially admire is that he combines scientific excellence with exceptional management. He creates an environment where ambitious work gets done thoughtfully and efficiently, and he makes something very difficult look almost effortless. That example has strongly shaped how I think about both research and leadership.
Dr. Lu is excited to join the UW-Madison after attending two other “UW” schools. “I was a postdoc at the University of Washington, began my faculty career at the University of Waterloo in Canada, and now I’m at the University of Wisconsin–Madison.” Dr. Lu looks forward to collaborative interactions as his new UW home.