Identifying the cellular hierarchy and drug response of AML using cytometric data

描述（由申请人提供）：多参数流式细胞术分析提供单细胞测量，这对于理解单个肿瘤内和肿瘤间的细胞异质性至关重要。尽管细胞术技术正在迅速发展，但分析复杂单细胞数据的方法仍然不足。现有的细胞术数据分析方法往往是主观的和劳动密集型的过程，需要用户对潜在的细胞表型有深刻的理解。这一限制已成为单细胞分析的关键瓶颈。我的长期目标是开发新的计算方法，使高维单细胞数据的客观分析成为可能。本应用程序的总体目标是开发和应用拓扑方法来客观地识别流式细胞术数据下的细胞层次结构，并推断AML患者细胞层次结构失调和药物反应。中心思想是将流式细胞术数据集视为高维细胞点云，并使用拓扑方法计算提取云的形状。根据初步数据，这种形状可以用来推断细胞异质群体的表型层次。具体目标是：(1)开发客观识别单个AML样本的细胞层次结构的方法；(2)利用AML患者细胞组成的变化建立治疗反应模型。提出的目标有望为单细胞数据提供新的分析方法，并为AML的分化和药物反应失调提供见解。提出的研究是创新的，因为它将细胞异质性视为具有分支的表型和功能变化的连续体，旨在在没有明确定义的细胞类型的先验知识的情况下识别这种连续体。提出的研究具有重要意义，因为它消除了高通量单细胞分析计算方面的瓶颈。本研究有望重塑细胞术数据的分析方式，促进细胞术分析在不同领域的细胞异质性研究。

DESCRIPTION (provided by applicant): Multiparametric flow cytometry analyses provide single-cell measurements that are critical for understanding the cellular heterogeneity both within individual tumors and across tumors. Even as the cytometry technology is rapidly advancing, approaches for analyzing complex single-cell data remain inadequate. The existing cytometry data analysis approaches are often subjective and labor-intensive processes that require users' deep understanding of the underlying cellular phenotypes. This limitation has become a critical bottleneck of single-cell analysis. My long-term goal is to develop novel computational approaches to enable objective analysis of high- -dimensional single-cell data. The overall objective of this application is to develop and apply topological methods to objectively identify the cellular hierarchy underlying flow cytometry data, and infer the dysregulation of cellular hierarchy and the drug response in patients with AML. The central idea is to consider a flow cytometry dataset as a high-dimensional point cloud of cells, and use topological methods to computationally extract the shape of the cloud. Based on the preliminary data, this shape can be used to infer the phenotypic hierarchy underlying heterogeneous populations of cells. The specific aims are (1) develop methods to objectively identify the cellular hierarchy underlying individual AML samples; (2) develop methods to model the treatment responses of patients with AML using the changes in their cellular compositions. The proposed aims are expected to produce novel analytical methods for single-cell data, and provide insight into the dysregulation of differentiation and drug response in AML. The proposed research is innovative, because it views cellular heterogeneity as a continuum of phenotypic and functional changes with branchings, and aims to identify this continuum without prior knowledge of well-defined cell types. The proposed research is significant because it removes the bottleneck in the computational aspect of high-throughput single-cell analysis. The proposed research is expected to reshape the way in which cytometry data are analyzed, and promote the use of cytometry analysis to study the cellular heterogeneity in diverse fields.