Our lab is focused on the development and application of computational toxicology models for environmental chemicals, nanoparticles and drugs in animals and humans.
We can develop many types of computational models, including physiologically based pharmacokinetic (PBPK) and quantitative structure-activity relationship (QSAR) models. To build robust models, we incorporate machine learning and artificial intelligence approaches into our models to develop machine learning-driven, AI-assisted or AI-enhanced models.
Our research is very exciting because our computational models can be used to address many research questions related to toxicology and environmental health. For example, our AI-assisted PBPK models can help support exposure and risk assessment of environmental chemicals and nanoparticles. Our machine learning-driven QSAR models can support high throughput screening of carcinogenicity and non-carcinogenic toxicities of thousands of compounds in animals and humans. Our AI-enhanced PBPK models can also support the design and safety assessment of new nanomedicines. Additionally, our PBPK models for drugs in food animals can help support human food safety assessment of animal-derived food products, such as meat, milk and eggs, and ultimately help protect food safety. From an animal welfare perspective, our computational models can serve as an alternative to traditional animal experimentation, which will help to replace, reduce and refine animal testing, thereby supporting the 21st century toxicity testing paradigm.
Our lab has had some successes in research, teaching, mentoring and services since we joined UF in 2021. In terms of research, we published the first AI-empowered PBPK model for nanoparticles in tumor-bearing mice. This project was supported by the National Institutes of Health. This model can help design new nanomedicines that are potentially more effective and safer. This model can also help reduce animal experimentation in preclinical trials of new nanomedicines. We also completed a machine learning-based QSAR model that can predict the plasma half-life of drugs in multiple food animals, including cattle, swine, sheep, goats, chickens and turkeys. This project was supported by the U.S. Department of Agriculture through the national Food Animal Residue Avoidance Databank program. This model can be used to estimate the withdrawal interval of drugs in food animals after extra-label uses, thereby helping protect safety of animal-derived food products, such as meat, milk and eggs.
Regarding teaching, our lab developed a new course, called Artificial Intelligence in Environmental and Global Health. This course has been included as a required or elected course for Department of Environmental and Global Health master’s and Ph.D. programs. We can proudly say that our future EGH graduates will be equipped with AI knowledge and skills after graduation. Besides teaching AI in environmental and global health at UF, I will also teach machine learning and AI in toxicology and environmental health at the national level by instructing continued education courses on this topic at the Society of Toxicology Annual Meeting in March 2025. Additionally, I am co-editing a new textbook, “Machine Learning and Artificial Intelligence in Toxicology and Environmental Health,” scheduled to be published in late 2025. This book will fill a scientific gap and provide a useful reference for future students to learn how to apply machine learning and AI technologies to study toxicology and environmental health sciences.
In the area of mentoring, since I joined UF, I have recruited three postdoctoral associates, four Ph.D. students and a few research assistants. I have also served as the faculty advisor for eight master’s students and as a dissertation committee member for 11 Ph.D. students at UF.
My goal is to maintain an extramurally funded research program in computational toxicology and environmental health. I hope to continue developing useful AI-empowered computational models to help address research questions related to nanomedicine, food safety, toxicology and environmental health. I also hope to have opportunities to mentor more students and trainees. I hope to share my knowledge and skills in computational toxicology and environmental health with students and trainees not only at UF, but also at the national and international levels.
UF has a world-class AI infrastructure, a dedicated AI supporting team, a variety of software resources and many training and collaboration opportunities, as well as strong institutional support on AI-related research and teaching. I feel UF is an ideal place to conduct AI-related research. I am deeply grateful to be part of the AI community at UF.