Search

Unprecedented technological advances in single-cell RNA-sequencing (scRNA-seq) technology have now made it possible to profile genome-wide expression in single cells at low cost and high throughput. There is substantial ongoing effort to use scRNA-seq measurements to identify the "cell types" that form components of a complex tissue, akin to taxonomizing species in ecology. Cell type classification from scRNA-seq data involves the application of computational tools rooted in dimensionality reduction and clustering, and statistical analysis to identify molecular signatures that are unique to each type. As datasets continue to grow in size and complexity, computational challenges abound, requiring analytical methods to be scalable, flexible, and robust. Moreover, careful consideration needs to be paid to experimental biases and statistical challenges that are unique to these measurements to avoid artifacts. This chapter introduces these topics in the context of cell-type identification, and outlines an instructive step-by-step example bioinformatic pipeline for researchers entering this field.

Single-nucleus RNA sequencing (sNuc-seq) profiles RNA from tissues that are preserved or cannot be dissociated, but it does not provide high throughput. Here, we develop DroNc-seq: massively parallel sNuc-seq with droplet technology. We profile 39,111 nuclei from mouse and human archived brain samples to demonstrate sensitive, efficient, and unbiased classification of cell types, paving the way for systematic charting of cell atlases.

Understanding how populations of human T cells leverage cellular heterogeneity, plasticity, and diversity to achieve a wide range of functional flexibility, particularly during dynamic processes such as development, differentiation, and antigenic response, is a core challenge that is well suited for single-cell analysis. Hypothesis-free evaluation of cellular states and subpopulations by transcriptional profiling of single T cells can identify relationships that may be obscured by targeted approaches such as FACS sorting on cell-surface antigens, or bulk expression analysis. While this approach is relevant to all cell types, it is of particular interest in the study of T cells for which classical phenotypic criteria are now viewed as insufficient for distinguishing different T cell subtypes and transitional states, and defining the changes associated with dysfunctional T cell states in autoimmunity and tumor-related exhaustion. This unit describes a protocol to generate single-cell transcriptomic libraries of human blood CD4+ and CD8+ T cells, and also introduces the basic bioinformatic steps to process the resulting sequence data for further computational analysis. We show how cellular subpopulations can be identified from transcriptional data, and derive characteristic gene expression signatures that distinguish these states. We believe single-cell RNA-seq is a powerful technique to study the cellular heterogeneity in complex tissues, a paradigm that will be of great value for the immune system.