Correcting the Mean-Variance Dependency for Differential Variability Testing Using Single-Cell RNA Sequencing Data

Nils Eling,Catalina A Vallejos,John C Marioni,Sylvia Richardson,Arianne C Richard

doi:10.1016/j.cels.2018.06.011

Abstract

SummaryCell-to-cell transcriptional variability in otherwise homogeneous cell populations plays an important role in tissue function and development. Single-cell RNA sequencing can characterize this variability in a transcriptome-wide manner. However, technical variation and the confounding between variability and mean expression estimates hinder meaningful comparison of expression variability between cell populations. To address this problem, we introduce an analysis approach that extends the BASiCS statistical framework to derive a residual measure of variability that is not confounded by mean expression. This includes a robust procedure for quantifying technical noise in experiments where technical spike-in molecules are not available. We illustrate how our method provides biological insight into the dynamics of cell-to-cell expression variability, highlighting a synchronization of biosynthetic machinery components in immune cells upon activation. In contrast to the uniform up-regulation of the biosynthetic machinery, CD4+ T cells show heterogeneous up-regulation of immune-related and lineage-defining genes during activation and differentiation.

Highlights

Heterogeneity in gene expression within a population of single cells can arise from a variety of factors
Increasing evidence suggests that this heterogeneity plays an important role in normal development (Chang et al, 2008) and that control of expression noise is important for tissue function (Bahar Halpern et al, 2015)
This includes the coefficient of variation (CV) (Brennecke et al, 2013) and entropy measures (Richard et al, 2016)

Summary

Introduction

Heterogeneity in gene expression within a population of single cells can arise from a variety of factors. In a seemingly homogeneous population of cells, the so-called unstructured expression heterogeneity can be linked to intrinsic or extrinsic noise (Elowitz et al, 2002). Single-cell RNA sequencing (scRNA-seq) generates transcriptional profiles of single cells, allowing the study of cellto-cell heterogeneity on a transcriptome-wide (Gru€n et al, 2014) and single gene level (Goolam et al, 2016). This technique can be used to study unstructured cell-to-cell variation in gene expression within and between homogeneous cell populations (i.e., where no distinct cell sub-types are present). A similar pattern occurs during gastrulation, where expression noise is high in the uncommitted inner cell mass compared to the committed epiblast and where an increase in heterogeneity is observed when cells exit the pluripotent state and form the uncommitted epiblast (Mohammed et al, 2017)

Methods

Results

Discussion

Conclusion