At the Francis Crick Institute, our lab studies how cells respond to perturbations by combining quantitative experiments with scientific machine learning (SciML) to discover fundamental principles of cellular function and self-organisation.

In the dry lab, we integrate mechanistic differential equation models, which embed domain knowledge, with machine learning techniques such as representation learning and neural network-based regression to enable data-driven discovery in biology. In the wet lab, we pair lab automation with multiplexed antibody assays to quantify single-cell responses to perturbations at scale. We integrate wet and dry approaches to investigate how protein dynamics–the signals they transduce and the cell states they define–are regulated across spatial scales, from the biophysics of individual proteins to cell–cell interactions, within diverse cellular processes, including developmental and mitogenic pathways.

We use this integrated approach to revisit long-standing concepts and theories in biology, such as allostery, mass-action kinetics and Waddington’s landscape, assessing their validity in the large-data limit and extending them to operate across scales and data modalities. In doing so, we discover new mechanisms and build models that are interpretable, predictive, and robust even in low-data regimes, enhancing our ability to understand and predict cellular behaviour in both physiological and pathological contexts.