CAREER: Robust and systematic molecular engineering with synthetic DNA neural networks and collective molecular robots

介绍和动机。生命中充满了编码在基因组中的复杂程序，它们协调分子进行感知、计算、反应和生长。为了合理地设计和合成具有与生命本身一样复杂的可编程行为的分子系统，该项目将采用一种以生物学为灵感和计算机科学为中心的方法。一方面，大自然非常成功地进化和选择了由简单的单个分子组成的最有效、最强大的生物程序。为了充分利用分子的潜力来创造复杂的可编程系统，我们需要借用生物学中的信息处理原理。另一方面，计算机科学在控制复杂性方面非常成功，已经制造出了数十亿个电子元件的设备。为了制造更复杂的分子装置，人们需要借用那些使计算机工程取得成功的概念。知识价值。如果成功，这个项目将改变人们对相互作用的分子网络可能表现出的行为以及如何合理地设计这些行为的理解的前沿。特别是，该项目将寻求以下问题的答案：(1)空间组织如何提高分子电路和机械的性能？受生物学中的空间组织（如神经线路和支架蛋白）的启发，该项目将开发在DNA纳米结构表面空间组织的分子系统，以构建更快更可靠的生化电路、高效的分子图灵机和具有复杂时空行为的DNA细胞自动机。(2)如何将类似大脑的从例子中学习的原理嵌入到相互作用的分子网络中？受大脑中学习和记忆形成规则的启发，该项目将探索简单有效的学习算法，以创建合成DNA神经网络，这些神经网络可以从其生化环境中学习并回忆生化信号的模式。(3)复杂的集体行为是如何从分子马达的简单个体行为中产生的？受蚂蚁觅食和白蚁群集等动物集体行为的启发，该项目将建造分子机器人，这些机器人作为个体具有简单功能，但作为群体执行复杂任务，如货物分类和迷宫解决。更广泛的影响。该项目将以一种与计算生物学和生物信息学中现有方法根本不同的方式将计算机科学带入分子科学领域：它不仅仅是使用计算机算法和程序来帮助设计和分析分子系统，而是更多地采用计算机科学原理来创建能够执行指令的生化系统，以执行分子水平上的任务。如果成功，该项目将有助于将计算机科学从传统的电子基板扩展到可以执行新的生物启发算法的新分子基板，并为化学和生物医学科学创造新的前沿，为智能自主化学合成和复杂疾病诊断和治疗提供潜在解决方案。该项目将通过设计创新课程，为学生和博士后提供教学和指导经验，支持本科生研究和学生竞赛，创造高度跨学科的研究环境，以及让更多女性参与科学，为教育和教育工作者的发展做出贡献。该项目还将通过开发开源软件工具、在YouTube上制作研究视频、举办公开讲座和采访科学电视连续剧，增加公众对科学和技术的参与。

Introduction and motivation. Life is full of amazingly sophisticated programs encoded in genomes, orchestrating molecules to sense, to compute, to respond, and to grow. Towards rationally designing and synthesizing molecular systems with programmable behaviors as sophisticated as life itself, this project will take an approach that is biologically inspired and computer science centered. On one hand, nature has been very successful in evolving and selecting the most efficient and powerful biological programs made of simple individual molecules. In order to use the full potential of molecules to create complex and programmable systems, one needs to borrow the information processing principles in biology. On the other hand, computer science has been very successful in mastering complexity, and devices with billions of electronic components have been manufactured. In order to build ever-more-complex molecular devices, one needs to borrow the concepts that enabled the success of computer engineering. Intellectual merit. If successful, this project will transform the frontiers of understanding about what possible behaviors a network of interacting molecules can exhibit and how one can rationally design such behaviors. In particular, the project will seek answers for the following questions: (1) How can spatial organization improve the performance of molecular circuitry and machinery? Inspired by spatial organization in biology such as neural wiring and scaffold proteins, the project will develop molecular systems that are spatially organized on the surface of DNA nanostructures, to build faster and more reliable biochemical circuits, efficient molecular Turing machines, and DNA cellular automata with complex spatial and temporal behaviors. (2) How can the brain-like principle of learning from examples be embedded within a network of interacting molecules? Inspired by the learning and memory-forming rules in the brain, the project will explore simple and efficient learning algorithms to create synthetic DNA neural networks that learn from their biochemical environment and recall patterns of biochemical signals. (3) How can sophisticated collective behaviors arise from simple individual behaviors of molecular motors? Inspired by collective behaviors in animals, such as ant foraging and termite clustering, the project will build molecular robots that have simple functions as individuals but as groups perform complex tasks such as cargo sorting and maze solving.Broader impact. This project will bring computer science into areas of molecular sciences in a way that is fundamentally different from the existing approaches in computational biology and bioinformatics: it is not just about using computer algorithms and programs to aid the design and analysis of molecular systems, but it is more about adapting the principles of computer science to create biochemical systems that can carry out instructions to perform tasks at the molecular level. If successful, this project will help extend computer science from the traditional electronic substrates to new molecular substrates that can execute new biologically-inspired algorithms, and create new frontiers in chemistry and biomedical sciences with potential solutions to smart autonomous chemical synthesis and complex disease diagnostics and therapeutics. The project will contribute to education and educator development by designing innovative courses, providing students and postdocs with teaching and mentoring experiences, supporting undergraduate research and student competitions, creating a highly interdisciplinary research environment, and involving more women in science. This project will also increase public engagement with science and technology by developing open-source software tools, producing research videos on YouTube, giving open lectures, and interviewing for science TV series.