ABI Innovation: Tools and databases for network-based plant systems biology with applications to understanding plant-virus interactions

该项目旨在改进制作小型芥菜植物拟南芥中相互作用分子网络模型的工具。为了证明这些方法的有效性，本研究中开发的资源将用于研究植物在几种条件下对病毒感染的免疫力。了解病毒如何以及何时克服植物防御，使我们能够控制美国南部作物的重要病原体，从而获得更高的产量和更少的农药使用等好处。虽然生物学家可以同时测量细胞中数万个分子的存在和数量，但了解它们是如何连接的，即导致功能的“网络”或“系统”，是一个更困难的问题。大量的测量必须得到适当的管理，以便它们是可用的，并且必须正确地添加额外的信息：重要的是要跟踪一种类型的分子的量如何随时间变化，而不是混淆不同的东西。同样重要的是要了解细胞的哪些部分相互影响：那些属于功能连接的通路（或网络），以及哪些是相互独立的。例如，植物有复杂的机制来保护自己免受生物和环境压力。信号通路引起植物的反应，它们受到内部遗传因素以及更容易观察到的外部因素的影响。如果细胞内参与传递和解释信号的每一种蛋白质（或其他分子）都是已知的，那么就有可能非常有效地预测最终的反应。然而，植物研究人员对生物体中的分子的了解并不像许多研究动物的研究人员那样多，这意味着有很多缺失的节点。这使得很难提出一个可以测试的具体预测：这项研究旨在克服选定植物途径的这个问题，展示当有足够的信息时可能发生的事情。该项目将积极吸引学生参与跨学科研究，特别注重招募代表性不足的群体。德克萨斯大学圣安东尼奥分校（UTSA）是一所西班牙裔服务机构。本项目的具体目标有四个：1）构建拟南芥全基因组转录调控网络，并在免疫应答基因中进行验证; 2）提高蛋白质间相互作用的预测和拟南芥防御子网络的鉴定; 3）对拟南芥免疫系统进行基于网络的分析，响应网络，以破译植物病毒RNA沉默抑制因子在植物免疫中的作用;和4）提供在线数据库和分析工具，为基于网络的植物系统生物学研究。该项目有望通过几个创新的想法显着改善基于网络的分析。首先，所提出的方法专注于通过为每个节点定义网络邻域并测试每个节点的邻域中的富集来提高单个基因的预测准确性。这与大多数现有的方法相反，这些方法对基因模块（物种内或跨物种）进行预测，因此缺乏对单个基因水平的质量控制。其次，结合蛋白质-蛋白质相互作用网络，基因共表达网络和样本-样本网络，这项研究提供了一个例子，在由全局转录组景观自动定义的动态背景下分析这些网络;因此，这项研究有望提供更具体的预测，可以通过实验验证。此外，在这项工作中，整合计算工具来表征防御相关的网络结构将显着提高研究任何数量的过程中的基因的共调节网络的作用的能力，包括但不限于植物和动物疾病，癌症或干细胞生物学，或基因表达的组织特异性中涉及的基因。该项目的结果可在http://cs.utsa.edu/~jruan/plantnet/上查阅。

This project aims to improve the tools for making models for networks of interacting molecules in the small mustard plant, Arabidopsis. To demonstrate the effectiveness of these methods, the resources developed in this research will be used to study the plant's immunity to a virus infection under several conditions. Understanding how and when the virus overcomes the plants defenses gives us a handle on controlling an important pathogen of crop plants in the Southern US, with benefits such as better yield and less pesticide use. While biologists can measure the presence and amounts of tens of thousands of molecules in a cell all at once, understanding how they are connected, the 'networks' or 'systems' that lead to function, is a much harder problem. The huge amounts of measurements have to be properly managed so they are usable, and additional information has to be correctly added: it is important to track how amounts of one type of molecule change over time and not mix up different things. It is also important to understand which parts of the cell affect one another: those that belong to a functionally connected pathway (or network), and which are independent of each other. For example, plants have complicated mechanisms to defend themselves against biological and environmental stresses. Signaling pathways cause the plant's response, and they are influenced by internal genetic factors as well as the external ones that are more easily observed. If every protein (or other molecule) inside the cell that plays a part in carrying and interpreting the signal is known then very effective predictions about the final response are possible. However, plant researchers don't know nearly as much about the molecules in their organisms as is available to many researchers studying animals, which means there are a lot of missing nodes. That makes it hard to come up with a specific prediction that can be tested: this research aims to overcome this problem for selected plant pathways, to showcase what is possible when there is sufficient information. This project will actively engage students in interdisciplinary research, with a particular focus on recruiting underrepresented groups. The University of Texas at San Antonio (UTSA) is a Hispanic-Serving Institution. This project has the following four specific aims: 1) to construct genome-wide transcriptional regulatory network in Arabidopsis with validation in immune-responsive genes; 2) to improve the prediction of protein-protein interactions and identification of defense subnetworks in Arabidopsis; 3) to perform network-based analysis of Arabidopsis immune-responsive network in order to decipher the role of plant viral RNA silencing suppressors in plant immunity; and 4) to provide online databases and analytic tools for network-based plant systems biology studies. This project promises to significantly improve network-based analysis with several innovative ideas. First, the proposed approaches focus on improving accuracy of predictions for individual genes by defining a network neighborhood for each node and testing for enrichment in the neighborhood for each node. This is in contrast to most existing approaches that make predictions on gene modules (within- or cross-species) and therefore lack quality control on an individual gene level. Second, combining protein-protein interaction network, gene co-expression network, and sample-sample network, this research provides an example to analyze such networks in a dynamic context automatically defined by the global transcriptomic landscape; as such, this study is expected to provide more specific predictions that can be experimentally tested. In addition, integration of computational tools to characterize defense-related network structure in this work will significantly improve the ability to study the role of co-regulated networks of genes in any number of processes, including but not limited to genes implicated in both plant and animal disease, cancer or stem cell biology, or tissue specificity of gene expression. The results of the project can be found at http://cs.utsa.edu/~jruan/plantnet/.