Co-expression in Single-Cell Analysis: Saving Grace or Original Sin?

• Published in: Trends in genetics Vol. 34; no. 11; pp. 823 - 831
• Main Authors: Crow, Megan, Gillis, Jesse
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 01.11.2018
• Subjects: Animals; cell type; co-expression; Computational Biology; Gene Expression Profiling - trends; Genome - genetics; Humans; replicability; Sequence Analysis, RNA - trends; Single-Cell Analysis - trends; single-cell RNA-seq; transcriptome; Transcriptome - genetics; co-expression; replicability; single-cell RNA-seq; cell type; transcriptome
• ISSN: 0168-9525
• DOI: 10.1016/j.tig.2018.07.007

As a fundamental unit of life, the cell has rightfully been the subject of intense investigation throughout the history of biology. Technical innovations now make it possible to assay cellular features at genomic scale, yielding breakthroughs in our understanding of the molecular organization of tissues, and even whole organisms. As these data accumulate we will soon be faced with a new challenge: making sense of the plethora of results. Early investigations into the replicability of cell type profiles inferred from single-cell RNA sequencing data have indicated that this is likely to be surprisingly straightforward due to consistent gene co-expression. In this opinion article we discuss the evidence for this claim and its implications for interpreting cell type-specific gene expression. Single-cell RNA sequencing approaches are vastly increasing in scale, with individual experiments routinely profiling thousands or even hundreds of thousands of cells. Despite technical limitations associated with low-input sequencing, cell classification through unsupervised clustering is surprisingly replicable across studies. This can be attributed to the intrinsic low dimensionality of cell types dominating the variability seen in expression profiles. Low dimensionality of expression profiles implies gene co-expression. An exploration of the history of co-expression highlights the perils of making gene-level inferences in light of collinearity, an issue that has previously arisen in cancer subtyping analysis. Co-expression has been both the saving grace and the original sin of single-cell RNA-seq, enabling sample characterization at the cost of gene-level inference.