Single-Cell Multiomics

Abstract

Single-cell RNA sequencing methods have led to improved understanding of the heterogeneity and transcriptomic states present in complex biological systems. Recently, the development of novel single-cell technologies for assaying additional modalities, specifically genomic, epigenomic, proteomic, and spatial data, allows for unprecedented insight into cellular biology. While certain technologies collect multiple measurements from the same cells simultaneously, even when modalities are separately assayed in different cells, we can apply novel computational methods to integrate these data. The application of computational integration methods to multimodal paired and unpaired data results in rich information about the identities of the cells present and the interactions between different levels of biology, such as between genetic variation and transcription. In this review, we both discuss the single-cell technologies for measuring these modalities and describe and characterize a variety of computational integration methods for combining the resulting data to leverage multimodal information toward greater biological insight.

Keywords: computation; integration; multimodal; multiomics; next-generation sequencing; single-cell.

Figure 1

An overview of multimodal technologies. Multimodal technologies listed involve the simultaneous measurement of transcriptomic, genomic, epigenetic, proteomic, or spatial information. Technologies listed are not comprehensive but represent many of the most prevalent technologies used, as described in more detail in the text. Abbreviations: CITE-seq, cellular indexing of transcriptomes and epitopes; DR-seq, genomic DNA–messenger RNA sequencing; FISH, fluorescence in situ hybridization; G&T-seq, genome and transcriptome sequencing; MERFISH, multiplexed error-robust FISH; Paired-Tag, parallel analysis of individual cells for RNA expression and DNA from targeted tagmentation by sequencing; PEA, proximity extension assay; REAP-seq, RNA expression and protein sequencing; scBS-seq, single-cell bisulfite sequencing; scM&T-seq, single-cell methylome and transcriptome sequencing; scRNA-seq, single-cell RNA sequencing; SHARE-seq, simultaneous high-throughput ATAC and RNA expression with sequencing; smFISH, single-molecule FISH; TEA-seq, simultaneous trimodal single-cell measurement of transcripts, epitopes, and chromatin accessibility. Figure adapted from images created with BioRender.com.

Figure 2

Experimental and data processing workflows of single-cell sequencing data. Typical processing involves tissue preparation, single-cell isolation, and sequencing (experimental steps are highlighted in green), followed by alignment, normalization, dimensionality reduction (DR), neighborhood graph generation, and cell clustering. This is followed by cell type annotation and downstream analysis. Steps where multimodal integration is often performed are highlighted in light blue and marked with an asterisk, with the integration phase (joint DR, late integration, or label transfer) listed underneath. Figure adapted from images created with BioRender.com.

Figure 3

Computational strategies for single-cell multiomic integration. (a) Types of input data: paired, unpaired, and partially paired. Different colors represent different modalities, and cell diagrams show whether these are measured on the same or separate cells. (b) Integration techniques: matrix decomposition, shared neighborhood graphs, joint clustering, manifold alignment, Bayesian methods, and neural networks. Figure adapted from images created with BioRender.com.