Michigan State University researchers are leading a seven-university team that has received an initial $4.6 million award from the National Institutes of Health to develop computational models aimed at improving how medications are developed and prescribed for women. The award could ultimately reach $12.8 million over three years.
The project brings together 13 researchers from Michigan State University, Rutgers University, Emory University, Tulane University, the University of Colorado Anschutz, the University of Michigan and the University of Utah.
The team is attempting to address a longstanding gap in medical research: clinical trials and drug-development processes often do not fully account for the significant hormonal changes women experience throughout their lives.
Hormone levels can change substantially during menstrual cycles, pregnancy, contraceptive use, menopause and hormone replacement therapy. Those shifts can influence how medicines move through the body and how patients respond to them, potentially affecting both efficacy and side effects.
The researchers plan to build advanced computational models capable of predicting how those hormonal changes influence drug behavior.
Lead investigator Teresa K. Woodruff, president emerita of Michigan State University and a Research Foundation Distinguished Professor, described the effort as a “moon shot” intended to integrate factors including age, reproductive stage, disease, drug effectiveness and potential side effects into a more complete model of women’s health.
The project is part of the NIH’s Computational Modeling of Hormone Homeostasis Initiative, a collaboration between the NIH Office of Research on Women’s Health and the Division of Program Coordination, Planning and Strategic Initiatives.
Across the national initiative, NIH is awarding $21 million to support research using computational modeling and human data to improve understanding of sex-specific hormonal biology.
A major focus of the Michigan State-led project will be metabolic health.
Conditions including obesity, type 2 diabetes, cholesterol disorders and thyroid conditions are widespread among women and can intersect with reproductive disorders.
Because reproductive hormones and metabolism influence one another, widely used treatments such as insulin and GLP-1 drugs may produce different effects depending on a patient’s hormone levels and reproductive stage.
The research team plans to use artificial intelligence, existing clinical information, computer modeling and lab-grown human tissues to create a more individualized framework for understanding those relationships.
One of the project’s first goals is to use AI to organize and digitize more than 40 years of hormone research into a freely accessible database.
That information will then help researchers construct a standard computational model of the 28-day menstrual cycle and its relationship with major metabolic organs including the liver, muscle and fat tissue.
The researchers will subsequently expand that model using real-world patient information.
The goal is to account for a broader range of biological conditions and life stages, including menopause, birth control use, diabetes and obesity.
Rather than treating women as a single uniform patient population, the model is intended to represent how biological and hormonal differences can influence medication response throughout an individual’s life.
The team will also test its computational predictions using three-dimensional human tissue models and lab-grown organoids involving tissues such as liver, muscle and ovaries.
Those biological models will provide another source of information for validating and improving the computer simulations.
Ultimately, the researchers intend to develop personalized treatment tools that could help clinicians determine more appropriate medication doses for individual patients.
Initial drug categories include metformin, insulin and GLP-1 medications.
The tools are intended to help physicians predict drug effectiveness while reducing the risk of harmful side effects associated with differences in hormone levels and metabolism.
That could have significant implications for precision medicine.
Rather than prescribing primarily according to population-wide averages, clinicians could eventually use computational models to account for biological variables such as age, reproductive stage, disease status and hormonal environment.
Michigan State said the resulting open-access platform is intended to provide healthcare providers with practical tools for anticipating medication effectiveness, identifying possible adverse effects and tailoring treatments more precisely for women.
The findings are also expected to inform NIH guidance, national safety standards and clinical protocols used when testing new therapies.
The project also addresses the fragmented nature of existing women’s health data.
Research on hormones, metabolism, disease and drug response has accumulated across decades, but much of that information resides in separate studies, databases and formats that can be difficult to combine.
The team plans to use AI and biomedical informatics to systematically extract, organize and standardize those datasets so they can be used for computational modeling.
The resulting infrastructure could give scientists a larger foundation for building models specifically around women’s biology rather than adapting models primarily derived from broader or male-dominated datasets.
All computational models and data generated through the NIH initiative are expected to be made freely available to researchers and healthcare professionals worldwide when the work is completed.
Researchers from multiple disciplines are contributing to the program.
The Michigan State participants include Woodruff, Sudin Bhattacharya, Brian Johnson, Rance Nault and Timothy Zacharewski.
The broader team includes Qiang Zhang of Emory; Shuo Xiao and Jiyang Zhang of Rutgers; Ariella Shikanov of the University of Michigan; Corrine Welt of the University of Utah; Mary Sammel and Nanette Santoro of the University of Colorado Anschutz; and Hao Zhu of Tulane.
The multidisciplinary structure reflects the complexity of the challenge.
The project combines reproductive biology, endocrinology, pharmacology, engineering, artificial intelligence, biomedical informatics and computational modeling to create a system capable of representing biological processes that change continuously throughout a woman’s life.
Researchers will also use new approach methodologies, or NAMs, including engineered human tissues that recreate changing hormone environments during ovulation, menstruation, pregnancy, menopause and hormone-related diseases.
Those laboratory models will be exposed to commonly used metabolic drugs, with the resulting responses feeding back into the computational models.
The $4.6 million initial NIH award gives the multi-university team resources to begin building that infrastructure, while the potential $12.8 million three-year award could support a much broader effort to turn decades of women’s health research into practical clinical tools.
If successful, the project could create a new framework for evaluating medications in women by treating hormonal variation as a fundamental component of precision medicine rather than an exception that must be addressed after a drug has already been developed.
KEY QUOTES:
“Developing computational models that provide insights into women’s health — including the consequences of disease, efficacy and side effects of medicines — and integrate age and reproductive cycle stage is a moon shot. Our world-class team is taking on this project to enable a generation of healthier women.”
Teresa K. Woodruff, Lead Investigator, President Emerita of Michigan State University and MSU Research Foundation Distinguished Professor
“Because female hormone levels are constantly shifting, precision medicine allows us to map out these complex interactions. This NIH-backed initiative will create the first computationally driven clinical tool designed to guide medical care across every stage of a woman’s life.”
Teresa K. Woodruff, Lead Investigator
“Empowered by AI, novel assays and legacy human data, we will develop mechanistically based computational models of female physiology that can make translational, quantitative predictions for women’s responses to metabolic therapies.”
Qiang Zhang, Associate Professor at Emory University’s Rollins School of Public Health
“Using state-of-the-art computational technology to examine how these changes interact with commonly used medications is a critical pathway toward supporting life-course women’s health.”
Nanette Santoro, E. Stewart Taylor Professor at the University of Colorado Anschutz and President of the Endocrine Society
