Single-cell multiomics measures DNA, RNA, and protein in the same individual cell rather than averaging signal across thousands of cells at once. That resolution uncovers rare clones, co-occurring mutations, and genotype-phenotype links that population-level methods cannot see, with growing use across oncology, hematology, and the quality control of cell and gene therapies.
For decades, most genomic and molecular profiling has relied on bulk methods: pooling thousands to millions of cells into a single tube and reporting one average signal across the mixture. That average answers many questions well, but it discards the identity of the individual cells that produced it. Two patients with the same average variant allele frequency (VAF) can have very different underlying cell populations, one dominated by a single expanding clone, the other a mix of many small clones with different treatment sensitivities. Single-cell analysis restores that resolution by measuring molecular features one cell at a time, then reconstructing the population from the individual measurements up.
What single-cell multiomics measures that bulk methods cannot
Single-cell multiomics refers to methods that capture more than one molecular layer, typically DNA, RNA, and protein, from the same individual cell rather than from separate aliquots of a bulk sample. Because each measurement stays linked to a single cell of origin, researchers can ask not just which mutations are present, but which mutations occur together in the same cell, and whether a given genotype corresponds to a distinct phenotype. A tumor with two subclones, each carrying a different resistance mutation, can look identical to bulk sequencing if the mutations occur at similar overall frequencies. Single-cell genotyping resolves them into two separate lineages with two separate clinical trajectories (Morita et al., 2020). Pairing genotype with single-cell protein assessment of surface markers, or with transcript expression, on the same cell adds a functional readout, connecting a cell's genetic identity to what it is actually doing.
Where single-cell resolution has already reshaped research
Oncology has been the proving ground for resolving clonal heterogeneity with single-cell multiomics. In acute myeloid leukemia (AML), single-cell DNA sequencing of nearly three-quarters of a million cells across more than 100 patients mapped the order in which driver mutations are acquired and showed that co-occurring mutations in signaling and epigenetic regulator genes shape how aggressively a clone expands (Morita et al., 2020). In myeloproliferative neoplasms (MPN), single-cell sequencing showed that patients carrying both a JAK2 mutation and a CALR or MPL mutation typically carry them in separate, independent clones rather than a single doubly mutated clone, a distinction bulk sequencing cannot make because it cannot assign two mutations to the same cell or to different cells (Thompson et al., 2021). In multiple myeloma, single-cell studies have traced how subclones carrying secondary mutations in genes such as NRAS, KRAS, and TP53 emerge and expand under treatment pressure, reshaping the tumor's genetic composition before relapse becomes clinically apparent (Li et al., 2025). Beyond cancer, single-cell approaches have become standard tools in immunology and developmental biology for reconstructing cell lineages and states from complex, heterogeneous tissue.
Can single-cell resolution move measurable residual disease detection into the clinic?
Measurable residual disease (MRD), the small population of cancer cells that persists after treatment and drives relapse, exposes two distinct limitations, not one. Bulk sequencing pools thousands to millions of cells before reporting a single averaged signal, so a small residual clone can be mathematically diluted below the assay's detection floor by the surrounding majority of normal cells. Flow cytometry does not share that problem: it is not a bulk method, and it already reads cells one at a time. Its limitation is different. It measures surface or intracellular protein markers without an attached genotype, so a truly malignant residual cell and a benign but immunophenotypically similar cell can look identical to the instrument even though they are genetically distinct. That phenotypic overlap, not signal averaging, contributes to relapse in a substantial share of patients classified as MRD-negative by current flow-based methods (Robinson et al., 2023).
Single-cell approaches that combine genotyping with immunophenotyping in the same cell address both limitations at once: they resolve clones cell by cell the way flow cytometry does, but attach a genotype to each cell the way flow cannot, and have reported sensitivity down to roughly 0.01 percent (Robinson et al., 2023). This kind of clone-resolved detection is being explored across AML, MPN, and multiple myeloma: distinguishing residual leukemic blasts from persistent clonal hematopoiesis in AML, tracking which clone is responding to a JAK inhibitor in MPN, and flagging an expanding resistant subclone earlier than a population-wide average in myeloma. None of these is yet a routine clinical test; the assays remain confined to research settings and require further validation before they inform treatment decisions.
Cell therapy quality control and gene therapy safety at single-cell resolution
Cell and gene therapies raise a version of the same resolution problem, but for manufacturing rather than diagnosis. A CAR-T cell product is not a single, uniform cell type: single-cell transcriptome profiling of clinical CAR-T products has shown that a manufacturing run contains multiple distinct T-cell subpopulations with different stimulation and exhaustion states, reproducible from donor to donor (Wang et al., 2021). Cell therapy quality control that relies only on bulk potency or phenotype averages can miss a shift in the balance of these subpopulations between manufacturing lots, even when the average readout looks unchanged.
Gene therapy safety raises a related question for products that use integrating vectors: where in the genome does the vector insert, and does any single insertion give one clone a growth advantage over others. Integration site analysis already tracks this at the population level, and it was integration-site data that first identified the mechanism behind leukemia cases in early retroviral gene therapy trials, where vector insertions near the LMO2 oncogene drove clonal expansion (Niederer & Bangham, 2014). Single-cell approaches extend this by linking a given integration site directly to the functional and proliferative state of the individual clone that carries it, rather than inferring clonal dynamics indirectly from a mixed population. That link is what a safety assessment ultimately needs: not just which sites are occupied, but which specific clones are expanding and why.
Frequently asked questions
What is the difference between single-cell sequencing and single-cell multiomics?
Single-cell sequencing typically measures one molecular layer, most often DNA or RNA, from individual cells. Single-cell multiomics captures two or more layers, such as DNA and protein or DNA and RNA, from the same individual cell, preserving the link between a cell's genotype and its measured phenotype (Morita et al., 2020).
Why can't bulk sequencing detect measurable residual disease as reliably?
Bulk sequencing reports an average signal across the entire sample, so a small residual clone can be mathematically diluted below the assay's detection floor by the surrounding majority of normal cells. Combined single-cell genotyping and immunophenotyping has reported sensitivities near 0.01 percent by resolving individual cells rather than averaging across them (Robinson et al., 2023).
Is flow cytometry a bulk method for measurable residual disease detection?
No. Flow cytometry measures individual cells one at a time and is not a bulk method. Its limitation for MRD is that it reads protein markers without an attached genotype, which can make a malignant residual cell and a benign, immunophenotypically similar cell look the same, a different failure mode than the signal dilution seen with bulk sequencing (Robinson et al., 2023).
Is single-cell MRD monitoring used in clinical practice today?
Not yet as a routine clinical test. Current single-cell MRD assays remain in the research and validation stage across AML, MPN, and multiple myeloma, and adoption into practice will depend on further clinical validation and regulatory review.
How does single-cell analysis apply to cell and gene therapy manufacturing?
For cell therapies, it profiles the subpopulation composition of a manufactured product at single-cell resolution for cell therapy quality control, which can reveal lot-to-lot differences a bulk potency average would miss (Wang et al., 2021). For gene therapy safety, it links a vector's integration site to the growth behavior of the individual clone carrying it (Niederer & Bangham, 2014).
A field converging on the same conclusion
Across oncology, hematology, and cell and gene therapy manufacturing, the pattern is consistent: a diluted bulk average or a phenotype read without a genotype can each obscure the signal a researcher or clinician needs most. Single-cell multiomics does not replace bulk methods or flow cytometry for every question, and most of its clinical applications remain in development rather than routine use. But the direction of the field, from clonal architecture in leukemia to comparability testing in CAR-T manufacturing, is toward measurements that stay resolved down to the individual cell.
World Single-Cell Day brings together scientists working across these applications to share how single-cell data is shaping their own research. Learn more about the range of work being presented and register at worldsinglecellday.com.
References
[1] Morita K, Wang F, Jahn K, et al. Clonal evolution of acute myeloid leukemia revealed by high-throughput single-cell genomics. Nat Commun. 2020;11(1):5327. DOI: 10.1038/s41467-020-19119-8
[2] Thompson ER, Nguyen T, Kankanige Y, et al. Clonal independence of JAK2 and CALR or MPL mutations in comutated myeloproliferative neoplasms demonstrated by single cell DNA sequencing. Haematologica. 2021;106(1):313-315. DOI: 10.3324/haematol.2020.260448
[3] Li S, Liu J, Peyton M, et al. Multiple Myeloma Insights from Single-Cell Analysis: Clonal Evolution, the Microenvironment, Therapy Evasion, and Clinical Implications. Cancers (Basel). 2025;17(4):653. DOI: 10.3390/cancers17040653
[4] Robinson TM, Bowman RL, Persaud S, et al. Single cell genotypic and phenotypic analysis of measurable residual disease in acute myeloid leukemia. Sci Adv. 2023;9(38):eadg0488. DOI: 10.1126/sciadv.adg0488
[5] Wang X, Peticone C, Kotsopoulou E, Gottgens B, Calero-Nieto FJ. Single-cell transcriptome analysis of CAR T-cell products reveals subpopulations, stimulation, and exhaustion signatures. Oncoimmunology. 2021;10(1):1866287. DOI: 10.1080/2162402X.2020.1866287
[6] Niederer HA, Bangham CRM. Integration Site and Clonal Expansion in Human Chronic Retroviral Infection and Gene Therapy. Viruses. 2014;6(11):4140-4164. DOI: 10.3390/v6114140
