Watching Cells Decide

Single-cell measurements give us exquisite snapshots, but the scientific question is rarely about the snapshot alone. The harder question is how a population of cells is moving through possibility: where it is flowing, where it lingers, and which futures remain open.

That is one reason I keep returning to dynamical systems language. Attractors, basins, and vector fields are not just mathematical ornaments. They offer a way to turn a pile of expression profiles into a map of commitment, hesitation, and transition.

field note

RNA velocity gives local direction; the harder work is stitching local arrows into a global picture that respects noise and biological plausibility.

RNA velocity is especially compelling here because it hints at local direction. The interesting work begins after that hint: how do we turn local arrows into a global picture that respects structure, noise, and biological plausibility?

I am interested in that middle layer between raw measurement and confident narrative. It is where geometry, machine learning, and biology negotiate with one another. Done well, the result is not merely a clustering of cells, but a more faithful story about how decisions unfold in living systems.