Statistical Methods for Gene Regulatory Analysis From Single Cell Genomics Data

单细胞基因组数据基因调控分析的统计方法

• 批准号: 10728206
• 负责人: Robert R. H Anholt
• 金额: $ 10.84万
• 依托单位: CLEMSON UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-10-06 至 2026-01-31
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10728206
• 关键词: Statistical; Methods; Gene; Regulatory; Analysis

Gene regulatory networks (GRNs) provide information on the cis-regulatory elements controlling contextspecific expression of target genes, as well as the transcription factors acting on these elements. Understanding the dynamics of gene regulation is fundamental for understanding how cells undergo specialization for different functions, despite having the same genome; how cells respond to different environments by modulating gene expression; and how non-coding genetic variants cause diseases. Inference of GRNs from genomics data is a systematic approach to study gene regulation. However, the accuracy of such inference is limited if the cellular context under interest is a heterogenous mixture. The development of single cell genomics technologies can fill this gap by providing high-resolution GRNs. Therefore, there is a compelling need for efficient statistical methods to infer GRNs from single cell genomics data. The long-term goal of this project is to obtain a mechanistic understanding of how noncoding genetic variants affect cellular context-dependent GRNs and influence phenotypes. Single cell transcriptomic (scRNA-seq) and chromatin accessibility (scATAC-seq) data provide information on different cellular features, i.e., gene expression and active regulatory element location, respectively. Integration of these two types of data will provide more accurate information on gene regulation. In Specific Aim 1, we will extend our initial studies inferring subpopulation-dependent GRNs from unpaired scRNA-seq and scATAC-seq data (supported by a COBRE in Human Genetics Pilot Project since 02/01/2022) by benchmarking existing methods for integrative analysis of unpaired scRNA-seq and scATAC-seq data to build an optimized pipeline for unpaired data analysis. We will develop a statistical method to infer subpopulation-specific GRNs and analyze large-scale published datasets to build a database of GRNs for hundreds of cellular contexts. In Specific Aim 2, we will develop statistical methods for comparative gene regulatory analysis based on single cell genomics data. The comparison of GRNs between samples from diseased versus healthy patients or between two different treatments is an important scientific problem. Thus, an efficient computational method for comparative gene regulatory analysis based on different types of single cell genomics data is needed. In Specific Aim 3, we will develop a method and software to infer cell type specific GRNs from sc-multiome data. This method and software would have a significant and broad impact by providing a detailed view of how trans- and cis-regulatory elements work together to affect gene expression in a cell type-specific manner.

基因调控网络(GRN)提供控制上下文特定的顺式调控元件的信息 目的基因的表达，以及作用于这些元件的转录因子。 了解基因调控的动力学是理解细胞如何经历 尽管拥有相同的基因组，但对不同功能的专门化；细胞如何对不同的 通过调节基因表达，以及非编码的遗传变异如何导致疾病。 从基因组学数据推断GRN是研究基因调控的一种系统方法。然而， 如果感兴趣的蜂窝上下文是异质混合物，则这种推断的准确性是有限的。这个 单细胞基因组学技术的发展可以通过提供高分辨率的GRN来填补这一空白。 因此，迫切需要有效的统计方法来从单个细胞中推断GRN 基因组学数据。这个项目的长期目标是机械地理解非编码是如何 遗传变异影响细胞上下文相关的GRN，并影响表型。单细胞 转录(scRNA-seq)和染色质可及性(scatac-seq)数据提供了不同 细胞功能，即基因表达和活性调控元件的定位。集成 这两种类型的数据将提供更准确的基因调控信息。在具体目标1中，我们 将扩展我们最初的研究，从未配对的scRNA-seq和 SCATAC-SEQ数据(自2022年2月1日起由人类遗传学试点项目COBRE提供支持) 对未配对的scRNA-seq和scatac-seq数据的综合分析的现有方法进行基准测试，以 为未配对的数据分析构建优化的管道。我们将开发一种统计方法来推断 亚群特定的GRN并分析大规模发布的数据集，以构建GRN数据库 数以百计的蜂窝环境。在具体目标2中，我们将开发比较基因的统计方法 基于单细胞基因组数据的调控分析。不同产地样品间GRN的比较 疾病患者与健康患者或两种不同治疗方法之间的比较是一个重要的科学问题。 因此，基于不同类型的比较基因调控分析是一种有效的计算方法 需要单细胞基因组学数据。在具体目标3中，我们将开发一种方法和软件来推断 来自sc-Multiome数据的特定于细胞类型的GRN。这种方法和软件将具有重要的和 通过提供跨法规和顺式法规要素如何协同工作来影响广泛影响的详细视图 以特定细胞类型的方式进行基因表达。