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

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

Thumbnail Image
Change log

Cell-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.

Publication Date
Online Publication Date
Acceptance Date
Bayesian, immune activation, single-cell RNA sequencing, statistics, transcriptional noise, variability, Animals, Biological Variation, Population, CD4-Positive T-Lymphocytes, Cell Differentiation, Cell Lineage, Computer Simulation, Gene Expression Regulation, Immunity, Lymphocyte Activation, Mice, Mice, Inbred C57BL, Models, Theoretical, Sequence Analysis, RNA, Single-Cell Analysis, Transcriptome
Journal Title
Cell Syst
Journal ISSN
Volume Title
Elsevier BV
Cancer Research UK (C14303/A17197)
Medical Research Council (MR/P014178/1)
NE was funded by the European Molecular Biology Laboratory (EMBL) international PhD programme. ACR was funded by the MRC Skills Development Fellowship (MR/P014178/1). SR was funded by MRC grant MC_UP_0801/1. JCM was funded by core support of Cancer Research UK and EMBL. CAV was funded by The Alan Turing Institute, EPSRC grant EP/N510129/1.