Core B: Data Management and Bioinformatics Core

核心 B：数据管理和生物信息学核心

• 批准号: 10207346
• 负责人: AVIV REGEV
• 金额: $ 20.02万
• 依托单位: HARVARD MEDICAL SCHOOL
• 依托单位国家: 美国
• 财政年份: 2017
• 资助国家: 美国
• 起止时间: 2017-07-05 至 2023-06-30
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10207346
• 关键词: Address; Adopted; Affect; Animal Model; Antigen-Antibody Complex; Bioinformatics; Biological Assay; Biology; Caring; Cell physiology; Cells; Classification; Cloud Computing; Collaborations; Complex; Computational Biology; Computer software; Computing Methodologies; Coupled; Data; Data Analyses; Data Set; Ensure; Experimental Designs; Follow-Up Studies; Foundations; Galaxy; Gap Junctions; Genes; Genetic; Genetic Screening; Genomics; Guidelines; Immune; Immune response; Immune system; Immunity; Immunologist; Immunology; Infrastructure; Intuition; Knowledge; Learning; Methods; Modeling; Molecular; Noise; Pathologic; Pathway Analysis; Pathway interactions; Patients; Phenotype; Physiological; Quality Control; RNA; Research Activity; Scheme; Software Tools; System; Systems Biology; Testing; Time; Tissues; Visualization; acute infection; cell type; complex data; computerized tools; data complexity; data management; data pipeline; data portal; data quality; experience; genetic manipulation; heterogenous data; in vivo; innovation; insight; large scale data; member; molecular modeling; novel; open source; programs; response; single cell analysis; single-cell RNA sequencing; tool; transcriptome; transcriptome sequencing; user-friendly

Reconstructing the circuits that control how immune cells adopt specific states and control an immune response is a major challenge, due to the diversity of cell types, the spectra of cell states, and their temporal changes over the course of the response. The combined advent of massive scale single cell genomics and large scale genetic CRIPSR screens suddenly provide an extraordinary opportunity to reconstruct a systems level model of the complex molecular and cellular processes that unfold during an immune response, including the cell types and states that compose the response, the regulators that control them, how cells affect each other, and how they integrate to form physiological and pathological responses. However, to learn such knowledge from massive, noisy and heterogeneous data there is an enormous need for sophisticated, innovative, robust, and scalable computational methods. These span include early data quality control and processing that addresses noise such as false negatives in single cell RNA-seq profiles, and the inference of the regulators that control cell types, states and temporal transitions, and guidelines for adaptive experimental design, from the number of cells to analyze to the choice of genes to perturb. In particular, because the current capacity to perturb genes in vivo is limited, ranking candidates for perturbation and refining their predictions and ranking as new perturbation data is collected is key for successful discoveries. Finally, because of the complexity of the data and of the underlying biology, achieving insights requires a close partnership between immunologists and computational experts. Unfortunately, successful methods and foundational datasets often remain out of reach for immunologists, absent software and data portals that would serve those. Here, we will leverage our extensive and pioneering expertise in computational biology for systems immunology – which we developed, harnessed and demonstrated in a long-term and close collaboration with the members of this Program – to develop and deploy computational methods and tools to bridge the gap between data and knowledge in systems immunology, and apply them in the context of the program’s projects. Specifically, we will develop, establish and maintain cutting-edge tools for the analysis of single cell RNA-seq data including identification of cell types, states, temporal transitions, and the associated pathways and signature, with high efficiency compatible with massive scale data (Aim 1). We will develop, establish and maintain cutting-edge tools to predict key regulators associated with these cell types, states and responses, as they unfold over time, and develop and use methods that rank regulators and select targets for genetic manipulation in CRISPR screens in vivo, followed by adaptive identification of new regulators following additional genetic screens (Aim 2). We will establish and maintain a public portal for all data, analyses and methods we collect, and release software tools as part of our open source Trinity package (Aim 3). Throughout, we will serve as an effective nexus for all project and core leads for all analysis purposes.

重建控制免疫细胞如何采取特定状态并控制免疫的电路 由于细胞类型的多样性，细胞状态的光谱及其时间特性， 在反应过程中发生变化。大规模单细胞基因组学和 大规模的基因CRIPSR筛选突然提供了一个非凡的机会， 在免疫反应期间展开的复杂分子和细胞过程的水平模型，包括 细胞类型和组成反应的状态，控制它们的调节器，细胞如何影响每个 以及它们如何整合形成生理和病理反应。然而，要学习这样的 从大量的、嘈杂的和异构的数据中获取知识， 创新的、稳健的和可扩展的计算方法。这些跨度包括早期数据质量控制， 处理，解决噪音，如假阴性单细胞RNA-seq概况，和推理， 控制细胞类型，状态和时间转换的调节器，以及适应性实验的指导方针 设计，从分析细胞的数量到选择基因进行干扰。特别是，由于目前 在体内干扰基因的能力是有限的，对干扰的候选者进行排名并改进其预测 并且当收集新的扰动数据时进行排序是成功发现的关键。最后，由于 由于数据和基础生物学的复杂性，实现洞察力需要以下方面的密切合作： 免疫学家和计算机专家。不幸的是，成功的方法和基础数据集通常 免疫学家仍然遥不可及，缺乏软件和数据门户网站，将服务于这些。 在这里，我们将利用我们在计算生物学方面的广泛和开拓性的专业知识， 免疫学-我们在长期和密切的合作中开发，利用和展示， 该计划的成员-开发和部署计算方法和工具，以弥合差距 系统免疫学中的数据和知识之间的关系，并将其应用于该计划项目的背景下。 具体来说，我们将开发，建立和维护用于单细胞RNA-seq分析的尖端工具。 数据，包括细胞类型、状态、时间转换和相关途径的识别， 签名，高效兼容海量数据（目标1）。我们将开发、建立和 保持尖端的工具来预测与这些细胞类型，状态和反应相关的关键调节因子， 它们随着时间的推移而展开，并开发和使用对调节因子进行排名和选择遗传靶点的方法。 在体内CRISPR筛选中进行操作，然后适应性地鉴定新的调节剂， 额外的基因筛选（目标2）。我们将建立和维护一个公共门户网站， 我们收集的方法，并发布软件工具作为我们的开源Trinity包（Aim 3）的一部分。 在整个过程中，我们将作为所有项目和核心线索的有效联系点，用于所有分析目的。