We measured the dynamics of decision-making by a minimal bistable gene network integrated in a synthetic cell model, free of external perturbations. Reducing the number of gene copies from 10⁵ to about 10 per cell revealed a transition from deterministic and slow computation to a fuzzy and rapid regime dominated by single-protein fluctuations. Fuzzy computation appeared at DNA and protein concentrations 100-fold lower than necessary in equilibrium, suggesting rate enhancement by co-expressional localization. Whereas the high-copy regime was characterized by a sharp transition, hysteresis and robust memory, the low-copy limit showed incipient strong fluctuations, switching between states, and a signature of cellular individuality across the decision-making point. Our work establishes synthetic cells operating rapidly at the single molecule level to integrate gene regulatory networks with metabolic pathways for sustained survival with low energetic cost.

中文翻译：

---

合成细胞中的决策：生物学计算的局限性

我们通过集成在合成细胞模型中的最小双稳态基因网络（无外部扰动）测量了决策的动态。从10 ⁵减少基因拷贝数到大约每个细胞10个分子，揭示了从确定性和缓慢的计算过渡到以单蛋白波动为主的模糊和快速变化的过程。在DNA和蛋白质浓度比平衡所需的浓度低100倍时出现模糊计算，表明通过共表达定位可以提高速率。高拷贝机制的特点是跃迁，滞后和鲁棒的记忆，而低拷贝极限则表现出初期的强烈波动，状态之间的切换以及整个决策点的细胞个性特征。我们的工作建立了在单分子水平上快速运行的合成细胞，以整合基因调控网络与代谢途径，从而以较低的能量成本实现了持续的生存。