![](/sites/default/files/styles/height/public/team/Jie%20Liu--Chair--Session%203.jpg?itok=mCZIrRu4)
Dr. Jie Liu received a Ph.D. in bioinformatics from the University of Arkansas at Little Rock. She then conducted postdoctoral research at FDA’s Center for Food Safety and Applied Nutrition and worked as a research scientist at Altamira, LLC. She joined FDA’s National Center for Toxicological Research (NCTR) in 2020 as a staff fellow in the Division of Bioinformatics and Biostatistics. Dr. Liu’s specialized research focuses on the development of machine learning models and databases for safety evaluation and risk assessment. She has developed toxicity databases and computational models for liver and other organ toxicity prediction by integrating data from multiple sources. Dr. Liu’s work also includes the development of machine learning models for in vivo toxicity prediction, opioid receptor binding activity prediction, and the construction of a cheminformatics system to manage the extractable and leachable chemicals for medical devices and their associated toxicity data.
OpenTox Virtual Conference 2023
Endocrine Activity Predictive Models
Endocrine activity is crucial for human health since endocrine pathways regulate growth, development, reproduction, metabolism, and tissue function. Endocrine disrupting chemicals are a group of chemicals may alter the endocrine system and disrupt endocrine function by mimicking endogenous hormones. Humans and wildlife are exposed to endocrine disrupting chemicals might alter endocrine functions through various mechanisms and lead to various adverse effects. Hence, it is important to identify endocrine disrupting chemicals for improving the public health and protecting the ecosystem. However, the experiments to identify potential endocrine disrupting chemicals is time consuming and expensive. Therefore, machine learning is an efficient and promising approach to screen and predict the potential endocrine disrupting chemicals.
In this session, our speakers will talk about the endocrine activity predictive models developed using various algorithms.