Pilot grants for novel applications of AI in cancer research (UF Health Cancer Center)

Pilot grants for novel applications of AI in cancer research (UF Health Cancer Center) is sponsored by UF Health Cancer Center. This program awards pilot funding to researchers at the UF Health Cancer Center to use artificial intelligence to advance innovative cancer research.

Pilot grants fund novel applications of AI in cancer research » Cancer Institute » UF Health Cancer Institute » University of Florida Two teams of UF Health Cancer Center researchers have been awarded pilot funding to use artificial intelligence to advance innovative cancer research.

One project will develop an AI-driven framework to identify cancer-specific RNA modification sites and investigate their role in tumor progression, with the goal of discovering novel biomarkers for cancer detection and therapeutic targets.

The other funded project will develop and evaluate an AI tool that simulates the behavior of a real patient, creating a more realistic simulation environment to enhance communication training for clinicians.

The AI pilots were supported by the Cancer AI Working Group , which aims to advance AI methods and applications in basic and translational cancer research, foster interdisciplinary collaboration, and support early-career investigators. Learn more about the research projects below. “AI-driven prediction of m7G RNA modifications in cancer” Rui Yin, Ph.

D. , left, and Mingyi Xie, Ph. D.

Assistant Professor, Department of Health Outcomes and Biomedical Informatics Associate Professor, Department of Biochemistry and Molecular Biology This project will take a transformative approach to studying RNA modifications in cancer by combining AI-based discovery with rigorous experimental validation.

RNA modifications, particularly N7-methylguanosine (m7G), play a crucial role in gene expression regulation, impacting RNA stability, translation and cellular function. Dysregulated m7G modifications have been increasingly linked to cancer development, yet the mechanisms are still not well explored.

Traditional experiments for identifying m7G sites are usually labor-intensive and costly, while existing computational models fail to capture essential spatial and contextual RNA dependencies. In the new project, the team plans to develop a foundational, biological large language model for accurate m7G site prediction.

The researchers will further identify cancer-specific m7G modifications linked to tumor development and validate their functional relevance through wet lab experiments. The results will provide new insights into the role of m7G in cancer, with the goal of identifying potential biomarkers and therapeutic targets.

“Developing AI-enabled experiential communication training to improve implementation of evidence based clinical practice: The SPARC prototype” From left: Carma Bylund, Ph. D. , Jason Arnold, Ed.

D. , and Stephanie Staras, Ph. D.

Professor and Associate Chair of Education, Department of Health Outcomes and Biomedical Informatics Senior Director, E-Learning, Technology and Communications Professor and Associate Chair of Faculty Development, Department of Health Outcomes and Biomedical Informatics This project aims to bridge the gap between research and practice with scalable, technology-driven training for clinicians.

The quality of communication directly influences critical outcomes in areas like HPV vaccination. While implementation strategies often integrate clinician communication training to address this gap, existing formats — typically delivered in-person or via Zoom — are resource intensive. To address this limitation, this project will develop a scalable, AI-enabled standardized patient avatar to support experiential communication training.

The innovation, called the Standardized Patient Avatar for Reflective Communication Practice, is designed to simulate patient interactions and deliver feedback based on best practices in clinical communication. The pilot study will focus on HPV vaccination, where physician recommendation is the strongest predictor of vaccine uptake. The research team has extensive experience and validated communication tools in this area.

The team will develop the avatar prototype by adapting an HPV vaccination training scenario and integrating AI-driven feedback mechanisms. Next, they will test the avatar with 10 pediatric clinicians at UF Health. If successful, the avatar will allow communication skills training to be widely disseminated.

That will support large-scale clinician engagement and enhance the delivery of cancer care, with the goal of improving patient outcomes. The UF Health Cancer Center’s pilot funding programs receive crucial support from the state of Florida through the Casey DeSantis Cancer Research Act ( Fla. Stat.

§ 381. 915 ).