The Office of Research on Women’s Health (ORWH), in partnership with the Division of Program Coordination, Planning, and Strategic Initiatives (DPCPSI) the Office of Research Innovation, Validation, and Application (ORIVA), and the Office of Strategic Coordination, is pleased to introduce the recipients of the Computational Modeling of Hormone Homeostasis awards. These awards will support the development of advanced mathematical models based on human data reflecting hormone activity across the life course, from adolescence to older age. Building on this investment, these models will improve the evaluation of new treatments and support more informed clinical decisions by strengthening researchers' ability to predict medication dosing needs and side effects from differences across individuals, improving care for people and communities. Beyond treatment evaluation, this initiative aims to advance understanding of the factors that shape health across populations by integrating consideration of biological differences into the study of hormone regulation. This allows for more robust and adaptable findings that result in stronger research studies, advance precision medicine, and better reflect real-world outcomes for women and men.
In FY26, six awards were made, totaling $21M. Learn more about the awardees and their compelling work below.
Data-Driven Computational Approaches for Hormone-Regulated Vaginal Tissue Remodeling Across the Lifespan (Virginia Polytechnic Institute & State University, PI: Raffaella De Vita)
The vagina is one of the body’s most hormone‑responsive organs, continually reshaping itself in response to hormones like estrogen and progesterone. Estrogen thickens tissue during puberty, hormones allow extensive stretching during pregnancy, and declining estrogen during menopause causes thinning and loss of elasticity. Yet conventional computer models treat vaginal tissue as static, leaving clinicians without tools to predict how hormone-driven changes influence injury risk or disease. This gap contributes to common long-term conditions, including pelvic organ prolapse, childbirth-related tearing, and postmenopausal vaginal atrophy.
Dr. De Vita’s team at Virginia Tech, University of Colorado Anschutz, University of California, San Diego, and University of Utah is developing a physics-informed computational model that simulates how vaginal tissue grows, remodels, and responds to mechanical stress across life stages. By integrating physical principles with biological data such as hormone levels, tissue stiffness, and genetics, the model will enable virtual clinical testing of surgeries, biomaterials, and hormone therapies tailored to an individual’s tissue characteristics.
Millions of women experience pelvic floor disorders and childbirth injuries, yet current treatments often rely on trial and error. This research will help clinicians anticipate severe tearing or pelvic floor muscle failure during childbirth and evaluate how a patient’s tissue may respond to hormone therapy or mechanical supports before treatment. Because nearly all soft tissues adapt to hormones, this project establishes a generalizable approach for studying hormone-driven degeneration and healing throughout the body. Physics-informed virtual organs could ultimately improve prediction, reduce reliance on animal models, and accelerate development of safer, more effective medical interventions.
Computational Modeling of Human Hormone Homeostasis in Bone to Predict Sex-Specific Disease and Therapeutic Responses (University of California San Francisco, PI: Edward Hsiao)
Hormones are key regulators of bone growth, maintenance, and repair, yet levels shift widely across life stages (puberty, pregnancy, menopause, aging) and during medical treatments such as hormone‑blocking therapies for breast or prostate cancer. These fluctuations can drive skeletal disorders, including osteoporosis and heterotopic ossification, but the biological links between hormonal change and bone disease remain poorly understood. Standard animal and lab models fail to capture human endocrine biology or the sex‑microbiome interactions essential for bone health. As a result, clinicians cannot reliably predict who will develop severe bone loss, how injuries will heal, or which drug will work best for an individual.
Dr. Hsiao and colleagues at University of California, San Francisco and San Francisco VA Medical Center are developing HERO, an AI‑enabled computational platform that models human bone biology by integrating clinical data, genetics, microbiome profiles, and biomechanical forces. HERO will simulate major hormonal transitions, such as menopause, age‑related testosterone decline, or cancer therapies, to predict bone loss and drug responses across virtual patient populations.
Osteoporosis disproportionately affects women; postmenopausal bone loss accounts for most fracture‑associated female deaths, and many breast‑cancer patients experience treatment‑related bone decline. HERO will help clinicians identify high‑risk patients and tailor protective therapies before fractures occur. More broadly, musculoskeletal disorders are a leading cause of disability worldwide. By modeling sex‑specific bone biology and hormone‑driven tissue responses, HERO can more accurately forecast disease trajectories, healing times, and treatment efficacy. The platform will also shed light on conditions without effective therapies, such as when bone tissue develops in soft tissues after surgeries or severe burns.
Modeling the Hormonal Influence on Central Metabolism: Drug Off-Target Prediction and Protein Synthesis (Montana State University Bozeman, PI: Ronald June)
We still lack a clear understanding of how sex hormones regulate cellular metabolism, and current metabolic models do not account for their effects. A flexible model system is needed to isolate hormone-driven changes at the tissue level. Without this, clinicians cannot accurately predict off‑target drug effects; for example, treatments for metabolically rooted diseases like osteoarthritis are prescribed without considering sex hormones, contributing to different responses in men and women.
Cartilage offers an ideal system for studying localized hormone‑metabolism interactions. The team will build a model that simulates cartilage‑specific metabolic pathways to predict sex‑specific drug effects. Their study will analyze chondrocytes from male and female donors, using the simple, nutrient‑limited cartilage environment to map metabolic behavior. Dr. June’s team at Montana State University, University of Texas Southwestern Medical Center, and University of Arizona will test cellular responses under varying hormone, nutrient, and oxygen conditions and validate model predictions against independent experiments. Ultimately, they aim to release a free, user‑friendly web tool to guide development of therapies tailored to women’s hormonal profiles and life stages.
Women experience higher rates and severity of osteoarthritis, especially after menopause. This project will generate new data on how cartilage cells use energy when estrogen and progesterone decline, providing critical data to inform hormone‑replacement strategies for musculoskeletal health. Because metabolic dysfunction underlies major chronic diseases, including cancer, Alzheimer’s, and diabetes, the model will help researchers forecast drug‑metabolism interactions, speed development of tissue‑repair therapies, and flag harmful side effects early. The simplified cartilage framework can later be expanded to study metabolic disease in more complex organs like the liver and brain.
Digital Twins of the Hypothalamus, placenta, and GI for the Therapeutics of Gestational Diabetes (Morgan State University, PI: Pilhwa Lee)
During pregnancy, the body must rewire its metabolism to supply the fetus with nutrients. Placental hormones signal the brain to induce temporary insulin and leptin resistance, raising maternal blood sugar and promoting fat storage. Yet the coordination between the brain and placenta—especially how the melanocortin system increases appetite and reduces energy expenditure—remains poorly understood. Without a clear map of this communication network, clinicians cannot reliably predict which women will develop gestational diabetes, excessive weight gain, or preeclampsia.
Dr. Lee and colleagues at Morgan State University and Johns Hopkins University are innovating a computational model that integrates the hypothalamus, placenta, and gastrointestinal tract to simulate pregnancy glucose regulation. Using longitudinal patient data and genetics, they will build personalized “digital twins” to test how hormonal and genetic differences drive vulnerability to gestational diabetes. These virtual patients will allow researchers to design optimized treatment strategies, including timing and dosing of insulin or metformin, to improve blood sugar stability.
Gestational diabetes and pregnancy‑related weight gain significantly increase the risk of lifelong metabolic disease. By clarifying the neural mechanisms behind these conditions, this work supports earlier interventions for high‑risk women, including those with obesity, which triples or quadruples the risk of gestational diabetes or preeclampsia. Digital‑twin–based therapy testing could reduce dangerous glucose swings, protecting both mother and fetus. Over time, the mathematical framework linking brain, gut, and hormonal systems may be extendable to metabolic disorders beyond pregnancy.
MOSAIC: A Mechanistic, Multi-Scale Modeling Framework for Hormone-Regulated Host-Microbiome Metabolism to predict Sex-Specific Therapeutic Efficacy and Toxicity (University of Virginia, PI: Jason Papin)
Medical treatments often overlook how biological sex influences drug processing in the body. Current technologies are limited in modeling how shifting hormone levels interact with human cells and the microbiome to influence drug effectiveness or toxicity. Existing computational models treat organs in isolation, leaving us unable to predict how hormone cycles, metabolism, and microbial communities jointly drive sex‑specific drug responses.
MOSAIC (Modeling Of Sex‑specific metabolism Across Interacting Communities) is a multi‑organ simulation that links mathematical models across tissues and includes a virtual microbiome to forecast bacterial responses to drugs. It will also model how estrogen is metabolized across gut, liver, and vaginal environments. Dr. Papin’s team at University of Virginia, University of Michigan, and University of Connecticut will apply MOSAIC to two hormone‑ and metabolism‑driven conditions: Bacterial Vaginosis (BV) and MASLD (Fatty Liver Disease), building an open‑source platform that can later expand to other hormone‑related diseases.
BV affects millions of women and behaves differently before and after menopause due to changing estrogen levels. MOSAIC will help explain antibiotic treatment failures and guide personalized recovery plans. Fatty liver disease also progresses differently in women, especially post‑menopause; MOSAIC will help tailor emerging metabolic therapies to female physiology. Clinicians will be able to predict drug metabolism based on sex and hormone profiles rather than relying on trial‑and‑error, while pharmaceutical developers can use MOSAIC to flag side effects or failures through virtual clinical trials before testing in humans.
Computational Modeling of Female Hormone Homeostasis and Drug Responses (Michigan State University, PI: Teresa K. Woodruff)
A woman’s hormone levels shift continually across the menstrual cycle, pregnancy, menopause, and birth control use. Yet women, especially those who are pregnant, have historically been understudied due to views of hormonal dynamics as overly complex. Standard endocrine models have also centered on male biology, studying female hormones mainly in reproductive contexts. In reality, these hormones influence many non‑reproductive processes, including metabolism. Conditions like Polyendocrine Metabolic Ovarian Syndrome (PMOS) show how tightly reproductive hormones and metabolism are linked. As a result, drugs with metabolic effects, such as insulin often behave differently in women, who may experience stronger side effects, altered drug clearance, or shifting dose needs across their lifespan. Lacking tools to more accurately predict these differences, clinicians struggle to tailor treatments, contributing to higher adverse drug reactions in women.
This project will build a computational model connecting female reproductive hormones with metabolic pathways to predict drug responses. Using artificial intelligence, the team will organize decades of clinical data and fill in gaps with experiments on lab‑grown human tissues. With this resource, Dr. Woodruff’s team at Michigan State University, Emory University, Tulane University, Rutgers University, University of Michigan, University of Utah, and University of Colorado Anschutz will construct baseline and disease‑specific models, particularly PMOS variants, and simulate how common metabolic drugs are absorbed, processed, and tolerated. The final product is a “virtual patient” platform that allows researchers to test drug performance under different hormonal conditions.
For decades, medications have not been optimized for female biology, leading to disproportionately more frequent side effects for women. This platform maps how shifting hormones influence drug metabolism, enabling clinicians to select safer therapies and doses whether a woman is using birth control, experiencing menopause, or managing PMOS. By creating an AI‑organized database of hormone–drug interactions, the project supports true precision medicine. These virtual modeling tools can later be adapted to test new classes of drugs, reducing reliance on long and costly clinical trials.
Modeling Hormone Biology with Precision: A Leap Forward for Women’s Health Research
This initiative marks an exciting step forward for women’s health research by combining human‑centered science with the power of quantitative approaches from math, engineering, and physics. By modeling hormone biology as it unfolds across the entire life course, and by treating sex as a biological variable, these projects open the door to clearer predictions, safer treatments, and more personalized care. Bringing together experts who understand sex differences in the human body with those who build advanced computational tools strengthens the science behind everyday health decisions. Together, these advances will help ensure that discoveries in hormone science translate into healthier futures for all communities, supporting ORWH’s mission to put science to work for the health of women and men.