Physical Biology: From Gene Regulation to Living Matter

# Active matter: when molecules build moving materials Watch the video first. Use this companion to revisit the reasoning and its evidence limits. Active matter is any system whose components consume energy locally to generate forces and motion: bird flocks, swimming bacteria, and inside cells, networks of filaments and motors. Active doesn't simply mean alive; purified proteins in a dish are active too, if fed fuel. And it's not perpetual motion: remove the fuel, and the motion stops. Here the filaments are microtubules: hollow tubes about twenty-five nanometers wide, built from tubulin. In cells they form transport tracks, and the spindle that separates chromosomes. The motors are proteins such as kinesin. Each kinesin step along a microtubule is powered by one ATP: about eight nanometers per step, pulling with a few piconewtons. No ATP, no walking. Now mix many microtubules, motors that can link to each other, and ATP. A linked motor pair can grab two filaments and walk on both, sliding them past each other. Across a network, the material flows, contracts and reorganizes. No. Each motor only walks on the filaments it touches. Large patterns, like star-shaped asters, emerge from many local interactions, with no central program. That is emergence. The Phillips Lab, with Matt Thomson's lab at Caltech, uses engineered kinesins that link only under light. Shine a pattern, and motors there link up and pull microtubules into a contracting network; the rest stays inert. In 2023, Banks and colleagues swapped in motors with different speeds and processivities, meaning how far a motor walks before letting go. Aster size depended on the motor. And in quasi one-dimensional experiments, contraction rate scaled linearly with single-motor speed. In a 2025 paper, Hirokawa and colleagues bleached a whole grid onto a contracting network. The bright squares moved inward at a rate set by motor speed, and also spread out, like diffusion but motor-driven, with an effective diffusion constant about a hundred times smaller than for free microtubules. To describe such materials, physicists coarse-grain filaments into continuous fields, such as density, orientation and flow, and write equations for how those fields change. A deforming grid directly tests the predicted flows. Collaborative work led by Fan Yang in Thomson's group, with the Phillips Lab, cut defects into light-made networks. V-shaped cracks merged below a critical opening angle and split open above it, a bifurcation reproduced by a continuum model. The team also used light and a predictive model to program micron-scale flows for transport, separation and mixing. Finally, the cost. A 2026 preprint by Duarte and colleagues, not yet peer reviewed, used a calibrated fluorescent ATP sensor as asters formed. It found ATP gradients over tens of microns, lasting tens of minutes, alongside motor-density gradients, and hypothesizes that maintaining the motor gradients accounts for much of the energy spent. ## Evidence guide EXPERIMENTAL MEASUREMENT: motor–microtubule assemblies and photobleached tracking measure motion in purified materials. COLLABORATIVE PRIMARY RESULT: geometry-dependent gap or crack merging is Thomson-led work, not topological-defect annihilation. CURRENT PREPRINT: measured energetic gradients motivate a motor-energy hypothesis. MODEL PREDICTION: continuum descriptions approximate collective behavior. Active matter requires local energy consumption; it is neither perpetual motion nor necessarily alive. Sources: [banks2023], [hirokawa2025], [yang2025prr], [duarte2026]. See the course bibliography and claim audit.