CORE 3: Modeling Core

核心3：建模核心 摘要 生物知识通常以分子网络的形式建模，分子网络是由基因和蛋白质组成的相互作用图。 基因或蛋白质-蛋白质成对相互作用。然而，生物系统并不是一个简单的大的成对的系统 网络，但由跨越生物学的生物子系统的深层和动态层次组成 比例。在这里，我们超越了基本的相互作用图，转而使用分子相互作用数据来直接推断 层次化子系统。这些计划是由称为网络提取的计算框架实现的 本体(Nexo)，我们最近展示的它能够捕获并实质上扩展已知的 基因本体论等途径数据库记录的细胞成分和过程的层次结构 (GO)。首先(目标1)，我们将分析分子网络上的生长数据来推断宿主-病原体基因 本体论，代表了对分子复合体和途径的全面、分层的描述 对寄主对病原体的反应很重要。这种分层结构将使用蛋白质- 项目1中产生的蛋白质相互作用数据，有公网数据支持；它将提供一个目标 通过系统地识别细胞的蛋白质模块及其相互关系来定义细胞。通过比较 这个数据派生的层次结构到文献精选的基因本体论(Aim 2)，我们可以识别新的子系统 对病原体有反应的生物。接下来，我们将使用此描述性层次结构来播种预测性全细胞模型。 利用深度神经网络的工具，遗传逻辑将被嵌入到细胞中的每个复合体/途径中 对这种结构的扰动如何引起宿主表型进行建模(目标3)。这个 神经网络结构将精确设置为AIM 1中组装的寄主-病原体基因本体的结构； 然后我们将训练这个神经网络，将项目2中的组合遗传扰动转化为 对宿主细胞反应的预测。这一层次不仅是描述性的，而且是预测性的，将 将细胞通路的基本知识转化为将这些知识用于治疗的框架。最后，在AIM中 4，我们将使用由核心生成的各种结构、生化、遗传和蛋白质组数据 确定寄主-病原蛋白复合体结构的综合建模方法。穿过 实现这些目标，我们希望大幅提高我们对结构和功能的认识 宿主对病原体的反应并为其提供最佳靶点的分子通路的层次结构 治疗性干预。

CORE 3: MODELING CORE SUMMARY Biological knowledge is often modeled in the form of molecular networks, interaction maps consisting of gene- gene or protein-protein pairwise interactions. Biological systems though are not simply one large pairwise network, but consist of a deep and dynamic hierarchy of biological subsystems ranging across biological scales. Here, we move beyond basic interaction maps to instead use molecular interaction data to directly infer hierarchical subsystems. These plans are enabled by a computational framework called Network-Extracted Ontologies (NeXO), which we have recently shown is able to capture and substantially extend the known hierarchy of cellular components and processes recorded by pathway databases such as the Gene Ontology (GO). First (Aim 1), we will analyze the growing data on molecular networks to infer a Host-Pathogen Gene Ontology, representing a comprehensive, hierarchical description of the molecular complexes and pathways important for the host’s response to pathogens. This hierarchical structure will be developed using the protein- protein interaction data generated in Project 1, backstopped by public network data; it will provide an objective definition of a cell by systematically identifying its protein modules and their interrelationships. By comparing this data-derived hierarchy to the literature-curated Gene Ontology (Aim 2), we can identify new subsystems that respond to pathogens. We will next use this descriptive hierarchy to seed predictive whole-cell models. Using the tools of deep neural networks, genetic logic will be embedded onto each complex/pathway in the cell hierarchical structure to model how perturbations to this structure give rise to host phenotypes (Aim 3). The neural network structure will be set exactly to that of the Host-Pathogen Gene Ontology assembled in Aim 1; we will then train this neural network to translate the combinatorial genetic perturbations from Project 2 into predictions of host cell responses. This hierarchy will be not only descriptive but also predictive, connecting basic knowledge of cellular pathways to a framework for using this knowledge therapeutically. Finally, in Aim 4, we will use various structural, biochemical, genetic and proteomic data generated by Cores using an integrative modeling approach for the structure determination of host-pathogen protein complexes. Through execution of these aims, we hope to substantially advance our knowledge of the structural and functional hierarchy of molecular pathways that host responses to pathogens and provide optimal targets for therapeutical intervention.