CPS:Synergy: Safety-Aware Cyber-Molecular Systems

网络分子系统是纳米级物理设备的网络，其中的计算控制着分子设备的行为和活动。在一些网络分子系统中，可编程控制器本身是由DNA门和算法实现的。网络分子系统未来计划的许多用途都是安全关键的，例如检测水中污染物或体内疾病标志物的生物传感器，以及仅将定制药物输送到体内检测到疾病的位置的药物疗法。问题是如何设计网络分子系统，使其在动态的、只被部分理解的物理环境中使用是安全的。该项目通过引入一种可以稳健实施的新的化学反应网络模型来解决这个问题。被模型接受的计算类别可以这样的方式实现，即使关键的物理参数被少量扰动，它们也将被证明正确地执行。该项目将利用这一进步，为创建网络分子系统建立端到端的安全意识视角。该方法需要紧密集成的计算和物理风险分析；基于反应网络的建模和验证技术的进步；以及对原型网络分子系统的模型驱动的实验和评估。该项目有可能使未来的网络分子系统变得更安全。RESULTS还将提供风险分析和自动验证能力，以满足未来对生物传感器和药物治疗等安全关键网络分子系统的认证需求。更广泛地说，生化反应网络的新模型将促进人们对如何从安全的角度分析在概率物理环境中运行的超大型纳米级网络物理系统的理解。该项目还将侧重于对不同学生群体进行跨学科培训，其成果将纳入本科生和研究生课程。

Cyber-molecular systems are networks of nanoscale physical devices in which computation controls the behavior and activities of the molecular devices. In some cyber-molecular systems the programmable controller itself is implemented by DNA gates and algorithms. Many of the planned future uses of cyber-molecular systems are safety-critical, such as biosensors to detect pollutants in water or disease markers in the body, and drug therapeutics to deliver a customized medicine only to the locations in the body where the disease has been detected. The problem is how to design cyber-molecular systems to be safe for use in a dynamic and only partially understood physical environment. This project addresses this problem by introducing a new model of chemical reaction networks that can be implemented robustly. The class of computations accepted by the model can be implemented in such a way that they will provably perform correctly even when crucial physical parameters are perturbed by small amounts. The project will leverage this advance to establish an end-to-end safety-aware perspective for creating cyber-molecular systems. The approach entails tightly integrated computational and physical risk analyses; advances in reaction-network-based modeling and verification techniques; and model-driven experimentation and evaluation on prototyped cyber-molecular systems. This project has the potential to make future cyber-molecular systems safer. Results also will provide capabilities in risk analysis and automated verification to meet the need for future certification of safety-critical cyber-molecular systems, such as biosensors and drug therapeutics. More broadly, the new model of biochemical reaction networks will advance understanding of how to analyze very large sets of nanoscale cyber-physical systems that operate in a probabilistic physical environment from a safety perspective. This project will also focus on interdisciplinary training of a diverse student population, and results will be integrated into undergraduate and graduate courses.