We study the principles of tissue organisation. By developing and integrating cutting-edge spatial genomics, artificial intelligence and gene editing tools, we build reference maps that reveal how changes in DNA and in the environment reshape the cellular circuits underlying phenotype. Tissues are where cells do their work, and where disease begins and progresses, so the quantitative data we generate offer a route to better treatments. Understanding tissue physiology also lets us design predictive models that recapitulate
disease in a dish.

We focus on conditions that affect women disproportionately, including gynaecological disorders and inflammatory and inflammatory diseases.

We find the reproductive system fascinating. It is a dynamic, ever-changing environment shaped by hormonal cycles, tissue-resident immune cells and age. The most striking example is the endometrium, the uterine mucosal lining, which sheds and then regenerates perfectly and without scarring — roughly 400 times over a lifetime.

This system is crucial for the perpetuation of the species and the transmission of genetic information to the next generation, and equally for maternal and fetal health and the lifelong health of the newborn. Defects in communication between the uterus and the placenta, a transient extraembryonic tissue, can profoundly alter developmental outcomes. Beyond reproduction, the hormones this system produces place it at the centre of women’s health more broadly. Understanding it comprehensively therefore matters regardless of whether a woman chooses to have children.

Reproductive disorders are a major global health burden, and one that is too often endured in silence. Around 10% of people born with a uterus are affected by endometriosis, and ovarian cancer remains one of the most aggressive cancers affecting people with ovaries.

The reproductive system also offers a uniquely powerful window onto general rules of tissue physiology. During menstruation the endometrium is destroyed and becomes inflammatory, then resolves that inflammation itself in order to regenerate without scarring — a rare example of inflammation resolving in an entirely physiological setting, and a model for how we might improve repair at other epithelial surfaces. In pregnancy, the same tissue becomes immunomodulatory, allowing embryo and mother to coexist despite a genetic mismatch that could otherwise trigger rejection — a natural setting in which to study how immune tolerance is induced, with relevance well beyond reproduction.

The complexity of the reproductive and immune systems demands a quantitative approach. Cells, as the fundamental unit of the human body, are the basis of our work. We operate within an iterative framework:

1. Profile: We are building a large, systematically curated collection of diagnostic tissue in which routine histopathology, spatial multiomics, whole-genome sequencing and
   clinical outcome data are all available for the same patient samples.
2. Predict: We use AI/ML to predict molecular and genetic state from tissue slides, and conversely to predict the histological consequences of a genetic variant.
3. Perturb: We use in vitro models to run large-scale perturbations, read out by single-cell techniques, to establish causality.

Each turn of this cycle integrates experimental data with human datasets, refining our computational models and improving their predictive accuracy.

The lab is based primarily at the Wellcome Sanger Institute and is highly multidisciplinary, bringing together biologists with a range of expertise alongside engineers and data scientists. Most of our team members work across both wet and dry lab, generating hypotheses from data and then testing them at the bench. We collaborate closely with other groups working on the rules of tissue physiology.

Roser Vento-Tormo is also a Global Adjunct Principal Investigator at AITHYRA, the Institute for Biomedical Artificial Intelligence. There her work focuses on developing AI methods for tissue biology in close collaboration with Alex Tong and Xinyi Zhang, with whom she co-supervises two PhD students. Shared interests include reconstructing 3D multiomics datasets from 2D tissue slides, methods to move from genotype to tissue phenotype, and novel modes of cell–cell communication.