Deciphering pro- and antiviral factors for CMV infection by heterogeneity sequencing

在过去的三年里，我们开发了单细胞SLAM-seq （scSLAM-seq）和生物信息学工具GRAND-SLAM来研究单细胞水平的实时转录反应。我们的方法能够在单细胞水平上进行剂量-反应分析。我们利用scSLAM-seq以前所未有的分辨率详细描述了小鼠巨细胞病毒（CMV）裂解成纤维细胞过程中的转录调控和潜在的分子机制。在这里，我们的目标是扩展计算方法，开拓基于高通量液滴测序方法的scSLAM-seq，并充分利用其在功能基因组学分析中的潜力。在过去的几个月里，我们将scSLAM-seq应用于商业化的10x Genomics Chromium平台。现在，这使得数千个单细胞的时间分析成为可能，并使这些技术得到广泛应用。因此，我们期望scSLAM-seq技术将广泛应用于许多研究领域。目前还没有可用的计算工具来分析数千个细胞的数据。我们将把GRAND-SLAM发展成为此类实验的标准工具。我们有质量控制的原型实现和对这些数据的探索性分析。我们将对它们进行扩展，将它们简化为一个R包，使它们公开可用。在这里，主要目标是启用可重复的分析，并识别在我们的经验中经常出现在此类实验中的陷阱。这对于本提案的目标以及其他研究人员进一步使用我们的技术是至关重要的。我们将发展“异质性测序”。scSLAM-seq允许将每个细胞在扰动前的状态与扰动后的状态联系起来。这允许利用“之前”单元的异质性，以便识别与“之后”任何参数的功能关系。我们将开发一个灵活的统计框架（并将其嵌入R包中）来执行这样的分析，同时考虑到单个单元数据的特性（稀疏性、退出、过度分散）。我们将使用“异质性测序”来鉴定与溶性巨细胞病毒感染相关的原基因和抗病毒基因。这包括在整体感染效率中发挥作用的因素，以及更细致的分析，以剖析对感染的特定阶段或细胞免疫反应的影响。候选药物将使用基因敲低和荧光报告病毒进行验证。

During the last three years, we developed single cell SLAM-seq (scSLAM-seq) and the bioinformatic tool GRAND-SLAM to study real-time transcriptional responses at single cell level. Our approach enables dose-response analysis at single cell level. We employed scSLAM-seq to detail transcriptional regulation and the underlying molecular mechanism during lytic murine cytomegalovirus (CMV) of fibroblast at unprecedented resolution. Here, we aim to extend the computational methods, pioneer scSLAM-seq for high-throughput droplet-sequencing-based approaches and exploit its full potential for functional genomics analysis.1. In the past months, we adapted scSLAM-seq for the commercial 10x Genomics Chromium platform. This now enables temporal analyses of thousands of single cells, and makes these techniques broadly applicable. We therefore expect that scSLAM-seq will be a widely used technique in many areas of research. At the moment there is no computational tool available to analyze such data for thousands of cells. We will develop GRAND-SLAM into the standard tool for such experiments.2. We have prototypical implementations for quality control and explorative analysis of such data. We will extend those, streamline them into an R package, making them publicly available and usable. Here, the main goal is to enable reproducible analyses and to identify pitfalls that, in our experience, often occur in such experiments. This is essential for the objectives in this proposal as well as for the further use of our technique by other researchers.3. We will develop "heterogeneity sequencing". scSLAM-seq allows to relate the state of each cell before perturbation to the state after perturbation. This allows to exploit the heterogeneity of the cells "before" in order to recognize functional relationships to any parameter "after". We will develop a flexible statistical framework (and embed it in the R package) to perform such analyses taking into account the peculiarities of single cell data (sparsity, drop-outs, over-dispersion).4. We will use "heterogeneity sequencing" to identify pro- and antiviral genes relevant for lytic CMV infection. This includes factors that play a role in overall infection efficiency, as well as more fine-grained analyses to dissect influences on specific phases of infection or the cellular immune response. Candidates will be validated using gene knockdowns and fluorescent reporter viruses.