A nationwide collaboration led by Michigan State University is funding MOSAIC, a computational modeling initiative to predict how a woman’s natural hormonal changes across life stages—from puberty and menstrual cycles to pregnancy, menopause, and hormone therapies—affect drug pharmacokinetics and pharmacodynamics.

The consortium, involving Rutgers, Emory, Tulane, University of Colorado Anschutz, University of Michigan and University of Utah, has secured about $4.6 million from the NIH with potential to reach $12.8 million over three years to advance women’s medication development and prescribing.

Funds will support creating sex-specific, data-driven models using interactive machine learning, multi-scale and AI-enhanced approaches, physics-informed modeling, and digital twins of biological systems.

The broader aim is to improve inclusive healthcare and health outcomes by accounting for sex-specific biology across the lifespan.

A core focus is treating sex as a biological variable (SABV) and adopting life-course perspectives to streamline drug testing and translate findings from bench to bedside.

The initiative seeks to translate SABV principles into innovative modeling to accelerate therapeutic development and clinical translation.

Leading voices emphasize SABV and lifespan perspectives, advocating for human-data-driven, mechanistic models rather than traditional animal models.

A central goal is a personalized microbiome digital twin that combines individual microbial data with AI-generated mechanistic equations to simulate biology and interventions.

The project aims to translate data into actionable models by linking hormone biology with metabolic therapies, addressing conditions such as obesity, type 2 diabetes, cholesterol disorders, thyroid issues, and reproductive syndromes like PMOS.

Researchers highlight the cross-disciplinary, open-data ethos and the potential to turn complex hormone–drug interactions into practical clinical decision tools.

MSU contributors include Sudin Bhattacharya, Brian Johnson, Rance Nault, and Timothy Zacharewski, with collaborators from Rutgers, Michigan, Tulane, Emory, Colorado and Utah.

Initial test cases will apply MOSAIC to bacterial vaginosis and MASLD, with plans to generalize the platform to broader therapeutic questions.