CAREER: Algorithms in nature: Uncovering principles of plant structure, growth, and adaptation

这个项目的目标是帮助算法成为描述生物系统使用的问题解决策略的首选语言。算法长期以来一直是计算的语言，所有生物系统都必须进行计算（即处理信息）才能生存。因此，对“自然中的算法”的研究可以代表计算机科学和生物学之间跨学科研究的一个新领域。更具体地说，该项目旨在揭示自然界分支结构共享的基本网络设计策略和优化原则，包括植物茎（地上）和根（地下）结构，以及大脑中的神经分支乔木。理解进化用来设计这些系统的基本模式有双向的好处；它可以让我们更好地理解这些自然网络是如何处理信息和在健康和疾病中发挥作用的，它还可以为构建更好的工程网络带来新的计算策略。在教育方面，将算法作为生命科学课程的一项要求，可以帮助培养下一代跨学科科学家。这个项目有三个目的。第一个目标是发现植物结构如何生长和适应不断变化的环境的原则。本研究将：(1)利用帕累托最优理论研究不同的网络优化权衡如何塑造植物枝结构的形状，以及不同的权衡目标如何根据环境和物种进行优先排序；(2)确定驱动优先级的分子机制（基因）；(3)确定植物芽使用什么搜索算法来寻找资源和制定生长策略。这些问题将使用3D激光扫描作物物种（番茄，烟草，高粱，玉米，水稻）生长在多个条件和时间点，以及具有不同遗传背景的模式物种来研究。总体而言，本研究将把计算机科学中通常研究的网络设计原理与驱动植物网络形成和适应的原理联系起来，并可能有助于设计和评估育种策略，以提高作物产量。第二个目标是建立网络战模型来研究植物间的竞争。该项目将：(1)创建“角斗士式”的竞技场，研究两种植物如何争夺有限的光线；(2)发展博弈论方法来评估是否会出现优势策略或稳定策略；(3)根据两种植物的种类和生长环境，量化竞争策略的差异。这一目标将导致植物社会互动的预测模型，可以根据哪些植物“相处”最好，为多文化养殖空间的设计提供信息。第三个目的是测试这些原理对其他生物和工程分支结构的普遍性。该目的将测试从植物茎结构中了解到的分支特性是否也决定了地下植物根结构和大脑中的神经（轴突和树突）结构的结构。例如，根架构也是帕累托最优的吗？他们用什么搜索算法来寻找营养物？竞争是如何影响这些策略的？这也将导致从植物到神经元这两个生命领域分支结构的首次定量比较。最后，该Aim还将研究人类工程网络，这些网络也必须根据动态环境中的资源可用性和需求调整其结构。生物学的见解可以揭示在战争或自然灾害后对受损基础设施进行最佳重建的新策略，或者将现有基础设施扩展到发展中地区。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The goal of this project is to help make algorithms a preferred language for describing problem-solving strategies used by biological systems. Algorithms have long been the language of computing, and all biological systems must compute (i.e., process information) to survive. Thus, the study of "algorithms in nature" can represent a new field of interdisciplinary research between computer science and biology. More specifically, this project seeks to uncover fundamental network design strategies and optimization principles shared by branching structures in nature, including plant shoot (above-ground) and root (below-ground) architectures, as well as neural branching arbors in the brain. Understanding the basic patterns that evolution has used to design these systems has bi-directional benefits; it can lead to improved understanding of how these natural networks process information and function in both health and disease, and it can lead to new computational strategies for building better engineered networks. Educationally, making the study of algorithms a requirement in life science curricula can help educate the next-generation of interdisciplinary scientists. This project has three Aims. The first Aim is to discover principles governing how plant architectures grow and adapt to changing environments. This Aim will: (1) study how different network optimization trade-offs sculpt the shape of plant shoot architectures using the theory of Pareto optimality, and how different trade-off objectives are prioritized depending on the environment and species; (2) determine the molecular mechanisms (genes) that drive prioritizations; and (3) determine what search algorithms are used by plant shoots to find resources and to strategize growth. These questions will be studied using 3D laser scanning of crop species (tomato, tobacco, sorghum, corn, rice) grown across multiple conditions and time-points, and of model species with different genetic backgrounds. Overall, this Aim will link network design principles commonly studied in computer science with those driving network formation and adaptation in plants, and may help design and evaluate breeding strategies to enhance crop yield. The second Aim is to develop models of network warfare to study plant-plant competition. This Aim will: (1) create "gladiator-style" arenas to study how two plants battle for limited light; (2) develop game theory methods to assess whether dominant or stable strategies emerge; and (3) quantify how competition strategies differ based on the species of the two plants and their growth environment. This Aim will lead to predictive models of plant social interactions that can inform the design of polyculture farming spaces based on which plants "get along" the best. The third Aim is to test the generality of these principles to other biological and engineered branching structures. This Aim will test if the branching properties learned from plant shoot architectures also dictate the structure of plant root architectures below ground and neural (axonal and dendritic) architectures in the brain. For example, are root architectures also Pareto optimal? What search algorithms do they use to find nutrients? How does competition affect these strategies? This will also lead to the first quantitative comparison of branching structures across two kingdoms of life, from plants to neurons. Finally, this Aim will also study human-engineered networks that also must adapt their structure to resource availability and demand in dynamic environments. Insights from biology could reveal new strategies for optimal reconstruction of damaged infrastructure after war or natural disasters, or extension of existing infrastructure into developing areas. URL: http://www.snl.salk.edu/~navlakha/This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.