The diverse landscape of modeling in single-cell biology
The diverse landscape of modeling in single-cell biology
复制标题
单细胞生物学建模的多样化前景
DOI：
10.1088/1478-3975/ac0b7f
复制
发表时间：
2021
期刊：
影响因子：
2
通讯作者：
Nie, Qing
中科院分区：
文献类型：
--
作者：
MacLean, Adam L;Nie, Qing
单细胞生物学必然是计算性的。所涉及数据集的大小和复杂性需要它。为了满足这一需求，人们正在开发广泛且不断扩大的计算方法和模型（图 1）。在本期特刊中，我们展示了一系列研究论文和评论，概述了单细胞生物学中定量方法和物理建模的现状。单细胞研究的历史与我们首次在显微镜下观察单细胞一样古老。然而，在过去十年左右的时间里，我们对单细胞进行定量测量的能力发生了一场革命。在测序技术进步的引领下，我们现在可以以前所未有的分辨率表征单细胞的特性，包括基因组、转录组、蛋白质组、表观遗传状态以及细胞从表面极性到趋化性的空间特性的测量。我们不仅可以单独分析这些状态，而且还能够将“多组学”数据集中的多种分析结合起来，以及将基因组测量与功能分析（例如实时成像或克隆追踪）相结合。单细胞RNA测序数据集规模的增长速度远远超过摩尔定律[1]。这导致了维数灾难，其解决方案依赖于数据处理的计算方法。此类数据集的计算数据分析的典型流程从数据标准化、可视化、细胞聚类和差异基因表达开始。然而，这些数据中还有更多尚未开发的洞察力。为了充分利用单细胞基因组学数据集的能力，我们不仅需要新的方法来从数据推断生物特性 [2, 3]，而且我们需要整合物理和动力系统的观点来构建单个细胞的预测模型，例如，当它们从中间状态 [4] 过渡到承诺状态时。
Single-cell biology is computational by necessity. The size and complexity of the datasets involved demand it. To meet the need, a broad and everexpanding range of computational methods and models are being developed (figure 1). In this special issue, we present a collection of research papers and reviews that offer an overview of the current landscape of quantitative methods and physical modeling in single-cell biology.Studies of single cells are as old as the microscopes under which we first visualized them. However, in the last decade or so, our ability to make quantitative measurements of single cells has undergone a revolution. Led by advances in sequencing technologies, we can now characterize properties of single cells in unprecedented resolution, including measurements of the genome, transcriptome, proteome, epigenetic states, and spatial properties of cells from surface polarity to chemotaxis. Not only can we assay these states individually, but we are able to combine multiple assays in ‘multi-omic’datasets, as well as combine genomic measurements with functional assays, such as live imaging or clonal tracking. The rate of growth of the size of single-cell RNA sequencing datasets far exceeds Moore’s law [1]. This leads to a curse of dimensionality, solutions to which rely on computational methods for data processing. Typical pipelines for computational data analysis of such datasets begin with data normalization, visualization, cell clustering, and differential gene expression. Yet, far greater insight can remain in these data untapped. To harness the full capacity of single-cell genomics datasets, not only do we need new methods to infer biological properties from data [2, 3], but we need to integrate physical and dynamical systems perspectives to build predictive models of individual cells, eg as they transition through intermediate states [4] towards commitment.
登录
查看更多内容
影响因子：
2
作者：
通讯作者：
--
影响因子：
2
作者：
Allen W. Zhang;Kieran R. Campbell
通讯作者：
Allen W. Zhang;Kieran R. Campbell
影响因子：
2
作者：
Wang, Qixuan;Wu, Hao
通讯作者：
Wu, Hao
影响因子：
2
作者：
Patrick S. Stumpf;F. Arai;B. MacArthur
通讯作者：
Patrick S. Stumpf;F. Arai;B. MacArthur
影响因子：
2
作者：
Tsai K;Britton S;Nematbakhsh A;Zandi R;Chen W;Alber M
通讯作者：
Alber M