In MarioniLab/compareSingleCell: Comparative Analyses of Single-Cell RNA-seq Data with Bioconductor

library(BiocStyle)
library(knitr)
opts_chunk$set(error=FALSE, message=FALSE, warning=FALSE, cache=TRUE)

Workflow version information

R version: r R.version.string

Bioconductor version: r BiocManager::version()

Package: r packageVersion("simpleSingleCell")

Overview

Single-cell RNA sequencing (scRNA-seq) is widely used to characterize the transcriptional heterogeneity of cell populations at single-cell resolution. A powerful approach is to use scRNA-seq in multi-condition experiments to detect changes in population composition and expression profiles between conditions. For example, a researcher could use this strategy to detect changes in cell type abundance after drug treatment [@richard2018tcell] or genetic modification [@scialdone2016resolving]. This provides more information about biology - and in particular, causality - than more conventional scRNA-seq experiments with only one biological condition.

The increased complexity of the experimental design is reflected in the increased complexity of the subsequent computational analysis. We need to consider how to merge samples from different conditions onto the same coordinate system; how to test for differences in abundance or expression between conditions, possibly in the presence of replicates; and how to interpret and visualize the results with condition-specific metadata. These must be considered in addition to low-level steps such as quality control and normalization, as covered in the r Biocpkg("simpleSingleCell") workflow [@lun2016step].

Here, we provide workflows for performing comparative analyses of multi-condition scRNA-seq experiments. Using droplet-based data [@zheng2017massively] from a study of the early embryo, we demonstrate key steps including sample merging and differential analyses. This is mainly performed with software from the open-source Bioconductor project [@huber2015orchestrating]. The aim is to provide readers with clear usage examples to facilitate construction of their own custom analysis pipelines.

Author information

Author contributions

A.T.L.L. developed and tested workflows on all datasets.

Competing interests

No competing interests were disclosed.

Grant information

A.T.L.L. was supported by core funding from Cancer Research UK (award no. A17197).

Acknowledgements

References

MarioniLab/compareSingleCell documentation built on May 25, 2019, 4 a.m.