Multiomics—and in particular single cell multiomics—is exciting a research community frustrated by bulk and single analyte analysis. The ability to overlay data from different, orthogonal modalities from the same cell, across thousands of cells, affords scientists a deeper characterization of cell types and states than is possible by interrogating a single modality alone.
When transcriptional data is overlayed with epigenetic data, for example, new gene regulatory interactions and networks can be discovered. Similarly, key expression markers found by RNA-seq can help to interpret chromatin accessibility profiles.
In addition, with a multiomic approach, the use of precious samples can be stretched to provide multiple readouts from the same cell.
Knowing which genes are being expressed in any given cell, and at what quantities, is crucial to understanding what that cell is, what it is doing, and how it differs from and fits in with other cells. When each cell is individually queried (as they are in single cell RNA-seq), it is possible to know which genes are being expressed alongside which others in that cell type, at the time and under the conditions the cells lived and were harvested. Examining the population in bulk, on the other hand, yields only an average of all the cells being queried, allowing key correlations to be missed. A simple example: half of a population expresses gene A and B at high levels, while the other half does not express A and B at all—making the average cell a medium expressor of both A and B.
Understanding the places in the genome that chromatin is accessible to the cell’s transcription apparatus gives insights into how gene expression is being regulated. Similarly, performing ATAC-Seq (assay for transposase-accessible chromatin using sequencing) on a single cell basis affords correlations of the open chromatin regions in any given cell, rather than providing an average for the population.
Combining single cell RNA-seq with single cell ATAC-seq allows for correlations to be explored not only between co-expressed genes, and between accessible chromatin sites, but between the genetics and epigenetics of the single cells as well.
The ability to link single cell transcriptomic and epigenomic analyses allows for a deeper characterization of cell types and states than does either single assay alone. Gene expression data, along with key canonical gene expression markers, lets researchers cluster cells into known cell types. Yet these cells may bin differently when clustered based on gene accessibility, because some genes—transcription factors, for example—may be expressed in cells where their binding sites are not accessible, and similarly, some cells may have accessible binding motifs for transcription factors that are not expressed. By superimposing the two data sets, not only can cell type (B or T cell, for example) be determined, but insights are gained into the cell’s state (resting or cycling, for example) as well.
New gene regulatory interactions can be discovered by linking open chromatin regions (from ATAC-seq data) and associated gene activity (from RNA-seq data) in individual cells, allowing for the identification and fine-tuning of cell type- (and even state-)specific gene regulation, without the need to make computational inferences from cluster annotation across data sets. Some cells that cluster together in the RNA-seq space—say, resting and cycling B cells—for example, may be very distinguishable sub-clusters in ATAC space. This granularity—which could not be achieved by looking at only a single ’omic’s data set—can allow for the teasing out of the genetic regulation leading to cell states and perhaps transitions between them.
Similarly, open chromatin landscapes can often be challenging to interpret without knowledge of which genes are being expressed. It may be difficult to say what sub-clusters of cells in ATAC space—the same resting and cycling B cell sub-clusters, for example—represent merely based on other accessible stretches of chromatin. Analyzing differential gene expression can help make sense of why the separation exists.
Samples—especially patient-derived samples—are often available in limited amounts and must be treated as precious commodities. Combining two platforms—in this case RNA-seq and ATAC-seq—into a single assay means that the researcher doesn’t need to choose between two essential data sets. Instead, they can get multiple readouts, extending the amount of data collected from that sample, while saving time and resources.
Single cell ’omics allows for an understanding of how readouts—in the present case, transcription of specific genes, and stretches of accessible chromatin—relate to each other, without inferences from bulk measurements. Combining multiple single cell ’omics into one platform allows for the synergy to learn more from less.