Academia Sinica Logo

Abstract

Active matter systems self-organize through local interactions among energy-consuming components. A key challenge is understanding how microscopic rules create robust macroscopic structures, such as the cell\'s cytoskeleton. Composed of filaments and molecular motors, the cytoskeleton\'s emergent architecture is vital for cell division, motility, and shape, yet the principles governing its formation are not fully known. Using agent-based simulations, we explicitly model motor-filament interactions to map the control parameters that define network architecture, including contractile networks, extensile bundles, and active gels. Remarkably, our simulations show that two antagonistic motor types can spontaneously organize into a polarity-sorted, spindle-like bipolar structure, the essential machinery for chromosome segregation. These results provide design principles for novel active materials and offer insights into motor protein diversity. Finally, I will discuss how this framework can be enhanced with machine learning to accelerate the discovery of models for cell motility and morphology.