Single-Cell Technologies Are Redefining Our Understanding of Human Development
Single-cell and spatial omics are revealing the hidden biology of early pregnancy and offering clues to treat complications.
Early pregnancy has long been described as a “black box” in human development. Key biological events that happen at this time determine whether a pregnancy will be successful or not. Our understanding of this extraordinary developmental journey has remained limited due to the difficulty of accessing the developing embryo (both practically and ethically) and the limited tools available for studying this process. As a result, common pregnancy complications like recurrent miscarriages and pre-eclampsia are still poorly understood, with limited treatment options.
A new revolution in single-cell technologies is changing this. Up until recently, researchers have characterized cells by their structure, function and location, but have been kept in the dark about their individual molecular profiles. This means that essential details, like the intricate set of instructions sent between cells, are missed. Single-cell technologies now allow researchers to isolate and profile each cell individually, capturing its unique molecular signature. This has opened up a whole new toolkit for understanding complex biological systems – such as piecing together the biological events of early pregnancy.
Dr. Roser Vento-Tormo, a group leader at the Sanger Wellcome Institute, heads a research team investigating women’s reproductive health and the developmental trajectory of early pregnancy using single-cell technologies. “The maternal–fetal interface embodies a delicate equilibrium between growth, invasion and immune tolerance – processes that are fascinating at a fundamental biological level but also central to many pregnancy complications,” explained Vento-Tormo. ‘‘The biology itself adds complexity and drawing robust conclusions requires larger sample sizes than in more static tissues, as well as experimental and computational technologies that enable us to disentangle this complexity.”
The single-cell analysis toolkit
The fundamental shift in sequencing technology was kick-started in 2009, when a pioneering study performed the first single-cell genome-wide mRNA sequencing on a mouse blastomere.1 This was largely enabled by the emergence of next-generation sequencing. The field has rapidly developed ever since, and today, single-cell RNA sequencing (scRNAseq) is a go-to tool for many researchers.
In a landmark study published in Nature in 2018, Vento-Tormo and her team used scRNAseq to investigate the early stage of pregnancy, where placental cells detach and migrate into the maternal uterus, forming branch-like structures that attach to the uterus.2 “These placental cells remodel the maternal arteries to ensure adequate nutrient delivery to the embryo,” explained Vento-Tormo. “For decades, we lacked a human-specific, cellular-resolution view of the maternal uterine–fetal placental interface to understand what ‘normal’ implantation and placentation looks like. Our 2018 single-cell atlas of the decidua and placenta provided that baseline and a window into how maternal and fetal cells interact.”
Figure 1. Single-cell sequencing can be used to measure the genome (scDNA-seq), the DNA-methylome or the transcriptome (scRNA-seq) of each cell from a population of cells. Credit: Technology Networks.
Link DNA Mutations to Gene Expression
This whitepaper explores how targeted single-cell assays can unify genotypic and transcriptomic data to expose clonal structures, track rare populations and reveal resistance mechanisms with unprecedented clarity.View Whitepaper
Spatial transcriptomics is also changing the landscape of single-cell technologies. Single-cell methods require cells to be dissociated first, losing information about where they are in relation to other cells. Huge advances in this field mean that it is now possible to keep this positional information by attaching unique spatial barcodes to the RNA, allowing researchers to map where different genes are being expressed within the tissue after sequencing.
In a recent paper published in Nature, Vento-Tormo and her team built on their previous research and used multiomics to define the full developmental trajectory of the placental cells migrating to and interacting with the uterus. They also used spatial transcriptomics to provide spatial context, mapping gene expression to its specific location in the tissue.4 “To understand what goes wrong in disease, we first need a clear reference for what ‘healthy’ looks like,” explained Vento-Tormo. “In our 2023 study, we were able to describe the microenvironments affecting cellular behavior across the full maternal-fetal interface, including the layer of smooth muscle under the endometrium, and how the immune cells have a key role in modulating placental invasion.”
Creating a human cell atlas
Vento-Tormo is also a key contributor to the Human Cell Atlas (HCA) – a global consortium of scientists with a mission to map every cell type in the human body. The rapid rise of single-cell and multiomics technologies has paved the way for this ambitious project. The HCA was co-founded in 2016 by Sarah Teichmann, professor of stem cell medicine at the University of Cambridge and former head of cellular genetics at the Sanger Institute.
“This ‘reference map’ allows us to better understand how healthy cells function and communicate, and how they work synergistically to form tissues and organs,” explained Teichmann. “As we add disease datasets into the atlas, we begin to better understand pathology and support development of new diagnostics and treatments.”
Formed of over 3,600 scientists globally, the HCA community has already made significant and impactful discoveries. Teichman’s research team recently published a paper in Nature, in collaboration with others from the HCA Heart BioNetwork, that sought to map the cells of the cardiac conduction system. The team made the unexpected finding that human pacemaker cells (which set the pace of our heartbeats) express a particular receptor that is the target of a gut hormone, and its synthetic analogues, which are found in diabetes drugs.5 As Teichmann explained, this supports a direct mechanism for increased heart rate when taking these medications, which was observed clinically but not explained at a molecular mechanistic level before.
This work shows how the HCA has already been instrumental in describing how different organs talk to each other, and how it can help to interpret the effects of drugs. The HCA is currently integrating this vast collection of data towards a first draft atlas of cells in the human body.
Big data
As the cell atlas moves into three dimensions by incorporating multiple layers of data from the different omics modalities, the role of computational tools can’t be underestimated. The shift from single-cell to multiomics technologies has created a data deluge – vast datasets that span a huge number of cells, with billions of potential data points. Advanced computational tools that can extract biologically meaningful data from this are essential. Machine learning (ML) tools can integrate high-dimensional data, learning the complex relationships between the different modalities. ML is also the primary tool for pattern recognition, like identifying cellular “neighborhoods” or microenvironments within a tissue or detecting anomalies in the data.
The role of computational biologists and bioinformaticians in the biology lab has never been more critical. To delve deeper into the developmental journey of placental cells, Vento-Tormo and her team developed a computational method to reconstruct the differentiation of these cells by combining the cells’ gene expression similarity and proximity in space. Marie Moullet is a PhD student working in Vento-Tormo’s group, and focuses on developing deep learning models for single-cell omics data. “This method allowed us to investigate branches and decision points during the differentiation process of the placenta cells,” Moullet said. “In general, machine learning helps us to learn structure from high-dimensional data and to standardize analyses across cohort, platform and laboratories.”
“Computation is at the heart of everything we do, from preprocessing and integration of multi-modal data to inference of cell-cell communication and spatial patterning,” explained Vento-Tormo.
Machine learning tools are also being increasingly developed for predictive analysis and are already used to help reconstruct dynamic processes, such as developmental pathways, or predict cell-cell interactions.6 As the field continues to develop, ML tools will shift towards a more predictive and simulative role, with deep learning models predicting clinical outcomes, such as drug efficacy and optimal dosage. This could have a significant clinical impact, shifting from reactive treatments (based on symptoms having already appeared) to a more predictive and preventive healthcare model, as ML models utilize cell data to forecast the risk of disease.7 “We are in an exciting time in this domain due to the opportunities from deep learning, generative AI and AI agents,” said Teichmann.
Optimize Natural Killer Cell Analysis with Spectral Flow Cytometry
This application note demonstrates how spectral flow cytometry overcomes these limitations with a 25-marker panel designed for high-resolution NK cell characterization.View App Note / Case Study
Looking ahead to the future of single-cell technologies
Looking towards the next decade, both Teichmann and Vento-Tormo predict that technologies will become cheaper, higher-throughput and more readily available. “I expect multiomics at cellular resolution on clinically sized cohorts will become routine,” said Vento-Tormo. Teichmann adds that access to large clinical sample sets will further accelerate understanding of disease pathways and treatment responses. Advances in computational approaches will also increasingly be used to optimize experimental designs, potentially refining in vitro model systems and making them more physiologically accurate.
As more computational tools and artificial intelligence/ML models become increasingly available, one of the challenges researchers will face is navigating this new, vast landscape of advanced technology. The challenge of choosing the “right” tool to analyze the data will shift the bottleneck of discovery from data generation to data interpretation. Some groups are already performing large-scale benchmarking studies to help create more standardization in the field.8,9
Clinically, the application of single-cell multiomics is still in its early stages, but discoveries from the HCA community have already provided new insights into cancer10, COVID-1911, lung diseases such as cystic fibrosis12, bowel disease13 and heart disease5 , among others. As cellular maps move into three dimensions, scientists can start addressing questions that have so far remained a mystery – like unravelling the biological mysteries of early pregnancy to shed light on common and devastating complications like pre-eclampsia and miscarriage.
Undoubtedly, single-cell technologies are not just a new technique, but an entire shift in how we can ask and answer questions about human biology.
References
1. Tang F, Barbacioru C, Wang Y, et al. mRNA-Seq whole-transcriptome analysis of a single cell. Nat Methods. 2009;6(5):377-382. doi: 10.1038/NMETH.1315;KWRD
2. Vento-Tormo R, Efremova M, Botting RA, et al. Single-cell reconstruction of the early maternal–fetal interface in humans. Nature. 2018;563(7731):347-353. doi: 10.1038/S41586-018-0698-6;TECHMETA
3. Lim J, Park C, Kim M, Kim H, Kim J, Lee DS. Advances in single-cell omics and multiomics for high-resolution molecular profiling. Exp Mol Med. 2024;56(3):515-526. doi: 10.1038/S12276-024-01186-2;TECHMETA
4. Arutyunyan A, Roberts K, Troulé K, et al. Spatial multiomics map of trophoblast development in early pregnancy. Nature. 2023;616(7955):143-151. doi: 10.1038/S41586-023-05869-0;TECHMETA
5. Kanemaru K, Cranley J, Muraro D, et al. Spatially resolved multiomics of human cardiac niches. Nature. 2023;619(7971):801-810. doi: 10.1038/S41586-023-06311-1;TECHMETA
6. Liu X, Zhang Z, Tan C, et al. Global trends in machine learning applications for single-cell transcriptomics research. Hereditas. 2025;162(1):1-20. doi: 10.1186/S41065-025-00528-Y/FIGURES/9
7. Zack M, Stupichev DN, Moore AJ, et al. Artificial intelligence and multi-omics in pharmacogenomics: A new era of precision medicine. Mayo Clin Proc Digit Heal. 2025;3(3):100246. doi: 10.1016/J.MCPDIG.2025.100246
8. Luecken MD, Büttner M, Chaichoompu K, et al. Benchmarking atlas-level data integration in single-cell genomics. Nat Methods 2021 191. 2021;19(1):41-50. doi: 10.1038/s41592-021-01336-8
9. Hu Y, Wan S, Luo Y, et al. Benchmarking algorithms for single-cell multi-omics prediction and integration. Nat Methods. 2024;21(11):2182-2194. doi: 10.1038/s41592-024-02429-w
10. Ganier C, Mazin P, Herrera-Oropez G, et al. Multiscale spatial mapping of cell populations across anatomical sites in healthy human skin and basal cell carcinoma. Proc Natl Acad Sci U S A. 2024;121(2):e2313326120. doi: 10.1073/PNAS.2313326120
11. Ziegler CGK, Allon SJ, Nyquist SK, et al. SARS-CoV-2 receptor ACE2 is an interferon-stimulated gene in human airway epithelial cells and is detected in specific cell subsets across tissues. Cell. 2020;181(5):1016-1035.e19. doi: 10.1016/J.CELL.2020.04.035
12. Sikkema L, Ramírez-Suástegui C, Strobl DC, et al. An integrated cell atlas of the lung in health and disease. Nat Med. 2023;29(6):1563-1577. doi: 10.1038/S41591-023-02327-2;SUBJMETA
13. Kong L, Pokatayev V, Lefkovith A, et al. The landscape of immune dysregulation in Crohn’s disease revealed through single-cell transcriptomic profiling in the ileum and colon. Immunity. 2023;56(2):444-458.e5. doi: 10.1016/j.immuni.2023.01.002
Add Technology Networks as a preferred Google source to see more of our trusted coverage.
