Mısırlı Göksel, Yang Bill, James Katherine, Wipat Anil
School of Computing and Mathematics, Keele University, Keele, ST5 5BG, U.K.
School of Computing, Newcastle University, Newcastle upon Tyne, NE4 5TG, U.K.
ACS Synth Biol. 2021 Dec 17;10(12):3304-3315. doi: 10.1021/acssynbio.1c00157. Epub 2021 Nov 11.
Engineering genetic regulatory circuits is key to the creation of biological applications that are responsive to environmental changes. Computational models can assist in understanding especially large and complex circuits for which manual analysis is infeasible, permitting a model-driven design process. However, there are still few tools that offer the ability to simulate the system under design. One of the reasons for this is the lack of accessible model repositories or libraries that cater to the modular composition of models of synthetic systems. Here, we present the second version of the Virtual Parts Repository, a framework to facilitate the model-driven design of genetic regulatory circuits, which provides reusable, modular, and composable models. The new framework is service-oriented, easier to use in computational workflows, and provides several new features and access methods. New features include supporting hierarchical designs via a graph-based repository or compatible remote repositories, enriching existing designs, and using designs provided in Synthetic Biology Open Language documents to derive system-scale and hierarchical Systems Biology Markup Language models. We also present a reaction-based modeling abstraction inspired by rule-based modeling techniques to facilitate scalable and modular modeling of complex and large designs. This modeling abstraction enhances the modeling capability of the framework, for example, to incorporate design patterns such as roadblocking, distributed deployment of genetic circuits using plasmids, and cellular resource dependency. The framework and the modeling abstraction presented in this paper allow computational design tools to take advantage of computational simulations and ultimately help facilitate more predictable applications.
设计基因调控电路是创建能够响应环境变化的生物应用的关键。计算模型有助于理解那些人工分析不可行的特别大型和复杂的电路，从而实现模型驱动的设计过程。然而，目前仍然很少有工具能够对正在设计的系统进行模拟。造成这种情况的原因之一是缺乏适用于合成系统模型模块化组合的可访问模型存储库或库。在此，我们展示虚拟部件存储库的第二个版本，这是一个促进基因调控电路模型驱动设计的框架，它提供可重复使用、模块化且可组合的模型。新框架是面向服务的，在计算工作流程中更易于使用，并提供了几个新功能和访问方法。新功能包括通过基于图形的存储库或兼容的远程存储库支持分层设计、丰富现有设计，以及使用合成生物学开放语言文档中提供的设计来推导系统规模和分层的系统生物学标记语言模型。我们还提出了一种受基于规则的建模技术启发的基于反应的建模抽象，以促进对复杂和大型设计进行可扩展和模块化建模。这种建模抽象增强了框架的建模能力，例如，能够纳入诸如阻碍、使用质粒进行基因电路的分布式部署以及细胞资源依赖性等设计模式。本文介绍的框架和建模抽象使计算设计工具能够利用计算模拟，最终有助于促进更具可预测性的应用。
