Computational Design of Nonlinear Functions using Nucleic Acids - Microsoft Research Voucher 16000070 - Engineering - Synthetic Biology

该项目旨在通过引入新的范例和技术来编程生物系统，并理解活细胞进行的计算，从而显著推进合成生物学领域。合成生物学确实是一场即将到来的伟大革命，但在可编程复杂性方面却未能很好地扩展。这种失败的一个原因是过度强调数字范式。在这个项目中，我们展示了如何通过推进最先进的DNA链位移（DSD）理论来利用DNA、RNA和酶合成可编程的混合动态电路来摆脱这一限制。我们的方法为计算核酸设备设计的统一框架奠定了基础，并有助于回答以前未探索的重要问题，如“给定的生物分子电路在湿实验室环境中可靠地运行多长时间？”和“给定的生物分子电路对湿实验室环境中外来细胞反应产生的干扰有多强？”利用我们的方法合成的计算核酸设备显示出巨大的潜力，可以实现广泛的生物技术应用，包括用于分子生物学研究的智能探针，复杂化合物的体外组装，高精度的体外疾病诊断，以及最终在活细胞内可编程的感知和响应系统。这种多样性的应用是由一系列的实施策略支持的，包括核酸链位移，定位到底物，以及使用具有聚合酶，nickase和外切酶功能的酶。然而，现有的计算设计工具无法以统一的方式解释这些策略。因此，我们还将我们的理论方法编码为逻辑编程语言，该语言允许设计和分析广泛的计算核酸系统。该语言扩展了标准的逻辑编程与一个新的方程理论，以表达核酸分子基序。它自动识别整个系统中存在的匹配主题，以便应用表示为逻辑规则或动态系统输出的指定转换。该语言具有足够的表达能力来编码具有复杂拓扑结构的核酸链位移系统的语义，以及由广泛的酶执行的计算，并且很容易扩展到新的实现策略。语言开发是与微软研究院（剑桥，英国）合作，并产生了软件“Visual DSD”，方便用户友好的硅设计这种生物分子电路-软件“Visual DSD”在Windows和MacOS平台上运行，并可以耦合到其他计算平台，如MATLAB和Python。合成这种生物分子电路的核心问题之一是动力学速率的选择。最近，麻省理工学院（Cambridge， MA）的Nielsen等人开发了一个自动设计这种电路的平台“Cello”。尽管如此，Cello的设计过程仅限于布尔电路，并没有利用人工智能和公共领域的元数据来提高电路功能的准确性和工作条件范围。我们展示了在这个项目中开发的新的生化见解如何帮助合成非布尔电路（如低通滤波器或比率计算或对数）和使用DNA/RNA/酶的混合系统，并显着提高湿实验室中无细胞蛋白质合成（CFPS）系统的产量。该项目的这种转化影响将通过与微软研究院（英国剑桥）和Arbor Bioscience（密歇根州安娜堡）的合作来实现。

This project aims to significantly advance the field of synthetic biology by introducing new paradigms and techniques for programming biological systems and for understanding the computations performed by living cells. Synthetic biology is truly a great revolution in the offing but has failed to scale well on programmable complexity. One reason for this failure has been the overemphasis on digital paradigms. In this project, we demonstrate how to get rid of this limitation by advancing the state-of-the-art in the DNA strand displacement (DSD) theory to synthesize programmable hybrid dynamical circuits using DNA, RNA, and enzymes. Our approach lays the foundation for a unifying framework for the design of computational nucleic acid devices and helps answer previously unexplored important questions such as "How long will a given biomolecular circuit perform reliably in the wet-lab settings?" and "How robust is a given biomolecular circuit to the interference created by the extraneous cellular reactions in the wet-lab settings?" The computational nucleic acid devices synthesized using our approach show great potential for enabling a broad range of biotechnology applications, including smart probes for molecular biology research, in vitro assembly of complex compounds, high-precision in vitro disease diagnosis and, ultimately, programmable sense-and-respond systems inside living cells. This diversity of applications is supported by a range of implementation strategies, including nucleic acid strand displacement, localization to substrates, and the use of enzymes with polymerase, nickase, and exonuclease functionality. However, existing computational design tools are unable to account for these strategies in a unified manner. Hence, we also code our theoretical approach a logic programming language that allows a broad range of computational nucleic acid systems to be designed and analyzed. The language extends standard logic programming with a novel equational theory to express nucleic acid molecular motifs. It automatically identifies matching motifs present in the full system, in order to apply a specified transformation expressed as a logical rule or as a dynamical system output. The language is sufficiently expressive to encode the semantics of nucleic strand displacement systems with complex topologies, together with computation performed by a broad range of enzymes, and is readily extensible to new implementation strategies. The language development is in collaboration with Microsoft Research (Cambridge, UK) and has resulted in the software "Visual DSD" that facilitates a user-friendly in silico design of such biomolecular circuits - the software "Visual DSD" runs on both Windows and MacOS platforms, and can be coupled to other computational platforms such as MATLAB and Python. One of the core problems in the synthesis of such biomolecular circuits is the choice of kinetic rates. Recently, Nielsen et al at the Massachusetts Institute of Technology (Cambridge, MA) have developed the platform "Cello" for an automated designed of such circuits. All the same, the design procedure of Cello is limited to only Boolean circuits and does not make use of artificial intelligence and public domain metadata to increase the accuracy and the range of operating conditions over which the circuits function as desired. We demonstrate how the novel biochemical insights developed in this project can help synthesize non-Boolean circuits (such as a low-pass filter or ratio computation or logarithm) and hybrid systems using DNA/RNA/enzymes and to dramatically increase the yield of cell-free protein synthesis (CFPS) systems in the wet-lab. Such a translational impact of the project will be achieved through collaborations with Microsoft Research (Cambridge, UK) and Arbor Bioscience (Ann Arbor, MI).