Computational analysis of cell-to-cell heterogeneity in single-cell RNA-sequencing data reveals hidden subpopulations of cells

Abstract

Recent technical developments have enabled the transcriptomes of hundreds of cells to be assayed in an unbiased manner, opening up the possibility that new subpopulations of cells can be found. However, the effects of potential confounding factors, such as the cell cycle, on the heterogeneity of gene expression and therefore on the ability to robustly identify subpopulations remain unclear. We present and validate a computational approach that uses latent variable models to account for such hidden factors. We show that our single-cell latent variable model (scLVM) allows the identification of otherwise undetectable subpopulations of cells that correspond to different stages during the differentiation of naive T cells into T helper 2 cells. Our approach can be used not only to identify cellular subpopulations but also to tease apart different sources of gene expression heterogeneity in single-cell transcriptomes.Single-cell measurements of gene expression, using imaging techniques such as RNA-FiSH (fluorescence in situ hybridization), have provided important insights into the kinetics of transcription and cell-to-cell variation in gene expression [1][2][3] . However, such approaches can examine the expression of only a small number of genes in each experiment, thus restricting our ability to examine co-expression patterns and to robustly identify subpopulations of cells. Protocols have been developed to overcome these limitations by amplifying small quantities of mRNA 4,5 , which, in combination with microfluidics approaches for isolating individual cells 6,7 , have been used to analyze the co-expression of tens to hundreds of genes in single cells 8,9 . These protocols also allow the entire transcriptome of large numbers of single cells to be assayed in an unbiased way. This was initially done using microarrays 10,11 but is more often now done using next-generation sequencing [12][13][14][15] . Such approaches have been used to model early embryogenesis in the mouse 16 and to investigate bimodality in gene expression patterns of differentiating immune cell types 17 .After the generation of single-cell RNA-sequencing (RNA-seq) profiles from hundreds of cells, one goal to identify subpopulations that share a common gene-expression profile. Some of these subpopulations may represent previously unidentified cell types. Additionally, by studying patterns of gene expression in different single cells, insights into the regulatory landscape of each cell population can be obtained.However, methods for identifying subpopulations of cells and modeling their gene regulatory landscapes are only now beginning to emerge 18,19 . To fully exploit single-cell RNA-seq data, we have to account for the random noise inherent to such data sets 20 and, equally important, to account for different hidden factors that might result in gene expression heterogeneity. Although the importance of accounting for unobserved factors is well established in bulk RNA-seq studies [21][22][23] , robust approaches to detect and account for confounding f...

APE1/Ref‐1 knockdown in pancreatic ductal adenocarcinoma – characterizing gene expression changes and identifying novel pathways using single‐cell RNA sequencing

“…Differences between cell batches were corrected by applying the scLVM r package (Buettner et al ., 2015). In conjunction with scLVM, the Biomart r package (Durinck et al ., 2009) was used to obtain a list of cell cycle‐annotated genes.…”

Modeling heterogeneity in the pluripotent state: A promising strategy for improving the efficiency and fidelity of stem cell differentiation

“…Recently, we have extended this rationale to predict lineage specifiers performing a differential network analysis on single‐cell gene expression data in different biological systems 74. In contrast to previous computational methods for analyzing single cell data for identifying lineage specifiers 80, 81, 82, which rely on the identification of trends in the expression of individual genes in different subpopulations, in this study we have demonstrated that information of the interactions among genes in the subpopulation‐specific GRNs, is key for predicting the genes triggering differentiation reported experimentally 74. These network models 73, 74, 102, 103 relying solely on transcriptomics data could be significantly improved by overlaying epigenetics data 25, 32, 33, for contextualizing regulatory interactions at the transcriptional level depending on the chromatin state – e.g.…”

“…For example, a data‐driven approach has been very useful for identifying switch‐like changes in expression of key regulatory factors, sequential waves of gene regulation, and expression of regulators triggering differentiation 80. A similar method has been developed for deriving single‐cell latent variable models (scLVM), which allows the identification of undetectable subpopulations of cells that correspond to different stages during the differentiation 81. Moreover, modeling gene expression changes in individual cells has proven to be effective for distinguishing cell subpopulations close to fate commitment, and for identifying putative regulators of commitment and probabilistic rules of transition between subpopulations 82.…”

“…These approaches will be helpful for generating models for differentiation of PSCs into specific cell types, pinpointing the signaling pathways, transcriptional factors, epigenetic regulators, and the cross‐talk among them, that could be targeted for improving the efficiency of cell‐type specific differentiation protocols. In comparison to data‐driven computational models 80, 81, 82, integrative networks models will provide key mechanistic insights of the regulation of heterogeneity in the pluripotent state, and a systems‐level picture of the interactions among key regulators triggering differentiation to specific cell fates, which will be essential for devising more effective differentiation protocols.…”

Single‐cell technologies to study the immune system

“…These computational methods become crucial for data interpretation because this new technology generates an incredible amount of data, which require faster and more standardized computational methods. The data are also ‘corrupted’ by numerous confounding factors and biases that need to be corrected for, using automated methods 16, 17, 18, 19, 20…”