AlphaMorph – A data-driven mechanistic framework to understand and predict morphogenesis

Symmetry breaking and morphogenetic transitions are ubiquitous features of development, observed across diverse biological systems. Yet whether general biophysical principles govern these complex processes across different contexts remains an open question. Here, we develop an interpretable and predictive data-driven framework grounded in physics-inspired morphological order parameters to quantify developmental dynamics. We apply this framework to a stem-cell-derived embryo-like system (stembryo). By constructing a morphological phase space from these quantitative metrics, we identify distinct morphological states and characterize their dynamical trajectories. Temporal alignment of these trajectories reveals bifurcation points in stembryo morphogenesis. We envision this framework to provide a path toward data-driven discovery of the organizing principles underlying morphogenesis.