By Eliza Dysart
The UF Health Cancer Institute’s fifth annual Cancer AI Symposium brought together researchers, clinicians, students and innovators on Oct. 22 to explore how artificial intelligence is changing cancer research and where the field may be headed.

Held at Malachowsky Hall, the symposium was structured to highlight emerging research and spark new opportunities for research collaborations across disciplines. Organized by the Cancer Institute’s Cancer AI Working Group, the symposium engaged researchers from across the university, from engineering to medicine, from veterinary medicine to agricultural sciences, and from law to biostatistics.
The event opened with remarks from Duane Mitchell, M.D., Ph.D., the institute’s director, and Ji-Hyun Lee, DrPH, associate director for cancer quantitative sciences, who co-leads the Cancer AI Working Group with Kiley Graim, Ph.D., and Ali Zarrinpar, M.D., Ph.D.
“We want to encourage translational research and create opportunities for new scientific discoveries,” said Lee.
She highlighted the three foundations of big progress: good data, powerful and scientifically sound methods, and computing infrastructure like hardware and technical resources. The fourth foundation — and what technology cannot replace — is people and human relationships. It’s about trust between collaborators, mentoring the next generation of scientists and exchanging different perspectives.
Simple things like hallway conversations can have a major impact on generating new collaborations, Lee said, underscoring the value of the symposium in providing a forum for researchers to exchange ideas and learn from people outside their field.
“Science has never been just about technology,” Lee said. “The more AI enables independent work, the more valuable collaboration becomes.”
Ultimately, patients are at the core of the work.
“In cancer research particularly, behind every data set is a patient,” Lee said. “Behind every question we ask and every AI approach we develop, there are patients and families hoping for better answers.”
Pranam Chatterjee, Ph.D., an assistant professor of bioengineering and computer and information science at the University of Pennsylvania, delivered the keynote address, titled “Designing Programmable Biologics with Generative Sequence Models.”

Chatterjee discussed how generative AI can be used to design peptides and other molecules that interact with disease-related proteins. His research aims to move beyond traditional drug development approaches by using AI models to generate potential therapeutics and optimize multiple characteristics at once, including binding ability, solubility, toxicity and half-life.
The symposium also looked beyond the scientific capabilities of AI to questions about how researchers can use these tools responsibly.
Barbara Evans, Ph.D., J.D., LL.M., a professor in the UF Levin College of Law, discussed the future of research and the importance of remaining compliant as researchers work with open-source information and emerging technologies. Her presentation highlighted the need to consider legal requirements, including Department of Justice compliance, alongside scientific innovation.
The event’s poster session gave students and researchers an opportunity to present projects applying AI and computational methods to cancer-related challenges.

Elizabeth Do, a researcher in the lab of Jatinder Lamba, Ph.D., in the UF College of Pharmacy, presented a project on a faster and potentially more cost-effective approach to genomic testing for patients with acute myeloid leukemia. Machine learning-based methylation classifiers were incorporated into the research to help analyze and classify patient data.
Ayana Price presented research on AI-driven structural modeling with molecular dynamics validation of mutation effects in SOCS1-JAK2 complexes. Her work used AI to model protein structures, identify mutations and evaluate how those changes could affect cancer-relevant signaling pathways.
“Working on the project also strengthened her interest in computational research,” said Price.
Jay Rosen, M.S., a research software engineer in the UF College of Education, presented the Simulated Patient Avatar for Reflective Communication, or SPARC, a web-based training platform that allows pediatric clinicians to practice conversations about the HPV vaccine with an AI-powered parent avatar.

Developed in collaboration with the UF College of Medicine, SPARC is designed to make communication training more accessible by allowing clinicians to practice skills such as empathy, listening and vaccine recommendations in a virtual setting. Rosen said 88% of pediatric clinicians in an initial pilot responded positively to the training. Researchers are looking at expanding the approach to additional scenarios, including oncology and tobacco cessation.
The Division of Population Health Sciences in the UF Department of Surgery used natural language processing to distinguish between chest CT scans ordered because of symptoms and those conducted as routine surveillance for patients who previously had non-small cell lung cancer. The project demonstrated how AI could perform a classification task that would otherwise require extensive manual review, potentially allowing researchers to work with larger patient datasets more efficiently.
Together, the keynote address, lightning talks and poster presentations reflected the symposium’s broader purpose of bringing different disciplines into the same conversation. Researchers demonstrated how AI can support drug discovery, genomic analysis, clinical communication and data interpretation, while other speakers addressed the legal and regulatory questions that accompany those advances.
The symposium concluded with poster awards and closing remarks, giving attendees another opportunity to connect across research areas and consider future collaborations in cancer and AI.
Congratulations to the poster winners:
- Derek Li, M.S.: “Digital Twins as Prognostic Information in Clinical Trials: Value, Robust Incorporation and Efficiency Gains”
- Jay Rosen, M.S.: “AI-enabled Simulated Practice for Scalable Clinician Communication Training: Development and Pilot Testing of SPARC”
- Keerthana Chinthala, M.S.: “A Multimodal Semantic Retrieval System Linking Pathology Reports, Whole-Slide Images and Model Attention for Veterinary and Comparative Oncology.”
