Our final foray into the world of single-cell analysis (with previous insights from Jonathan Sweedler, Andy Ewing and Jim Eberwine) takes us to the cutting edge of spatial omics—a field that is rapidly reshaping our understanding of tissue biology. As a field, spatial biology has the potential to transform our understanding of disease by showing how cells behave in situ, within the full architecture of the tissue microenvironment.
Here, we speak with Todd Dickinson, CEO of Stellaromics, whose new platform brings true 3D spatial resolution to multi-omic analysis. He discusses the technical breakthroughs behind the system, the challenges of going from two to three dimensions, and what this shift could mean for the future of biological discovery.
Could you give me a general introduction to spatial omics as a field?
When introducing the field of spatial omics, I like to place it within its historical context.. Each decade has brought a key technological breakthrough that helped move the field forward. In the 1990s, it was the microarray, which enabled massively parallel experimentation on a single chip, allowing researchers to measure gene expression across thousands of genes and multiple samples simultaneously. One challenge with microarrays, however, is that you need prior knowledge to determine what you’re looking for and to pre-apply it to the array.
The 2000s saw the advent of next-generation sequencing (NGS), which enabled unbiased, non-targeted detection, allowing for the sequencing of any component in a sample without prior knowledge. NGS transformed genomics and has had an enormous impact across the life sciences, but it, too, has its limitations. With NGS, you’re essentially blending your entire sample – sequencing a homogenized mixture – and so all context regarding the distribution and location of specific genes or cells within the tissue is lost.
In the 2010s, the field addressed the issue of bulk measurement by enabling the analysis of heterogeneity of individual cells within a sample. While this represented a significant advancement, the critical spatial context of cells was still lost in single-cell sequencing. You could see which cells were there and what they were expressing, but you didn’t know where those cells were within the tissue or how they were organized relative to one another.
In the 2020s, we have witnessed the rise of spatial biology, especially in spatial transcriptomics. This field developed from the realization that to truly understand biology, it’s not enough to know which genes or cells are present – you also need to know their relative location. Spatial technologies allow us to map gene expression within the tissue's architecture, connecting molecular data to its physical structure. The resulting insights from this new capability have the potential to be truly transformative, leading to spatial transcriptomics being named Method of the Year by Nature in 2020. But to realize the true potential of spatial transcriptomics, it will be critical to push the field beyond thin 2D slices to richer, three-dimensional data that capture a more accurate picture of biology. This is our mission at Stellaromics.
Could you explain how your new 3D spatial multi-omics approach works?
Our focus at Stellaromics – specifically with the Pyxa™ platform – is to bring spatial multi-omics into the third dimension. Suppose you look at the technologies launched in the early part of this decade by other companies in the field. In each case, they’re trying to tackle a difficult task: capturing three-dimensional biological information by assembling data from multiple thin, two-dimensional tissue slices, typically 5-10 microns thick. Cutting through tissue in this manner destroys a significant number of cells. Inevitably, this results in the loss of a substantial amount of information, including cell morphology and the vertical relationships between cells.
Stellaromics’ Pyxa platform delivers a significant increase in the amount of information scientists can access by visualizing thick tissue sections of up to 100 microns — approximately 20 times thicker than the standard. The result is the ability to generate stunning 3D images and high-plex multi-omic data packages of tissue at subcellular resolution. We're not just identifying cells; we can pinpoint the precise location of transcripts within the cells, known as subcellular localization.
Three-dimensional spatial multi-omics allows for numerous advances in our understanding of cell and tissue morphology. It enables researchers to observe intact cellular layers instead of a flattened, disrupted monolayer. This opens a new realm of biological exploration, allowing you to maintain full cell morphology and detect rare events that might otherwise be missed in a thin slice. For example, you might observe small neuronal clusters that would be invisible if the slice wasn’t made in precisely the right place.
You also gain a more comprehensive view of cell-cell interactions, as you can capture their full spatial relationships in three dimensions. It becomes possible to visualize complex structures, such as vasculature, and understand how different cell types cluster around them. Overall, our 3D spatial multi-omics approach aims to preserve biological complexity, unlocking a level of resolution and context that is not possible in 2D.
Were there any analytical breakthroughs or major hurdles that your team had to overcome during development?
To image thick tissue, we first needed a way to make it optically transparent – to "clear" the tissue so we can image deep into it. Our founder, Karl Deisseroth, invented a tissue-clearing method called CLARITY, which is now widely used worldwide. That gave us a solid foundation of expertise in tissue clearing that we could apply directly to this challenge.
Even after the tissue is cleared, you still must image through a significant volume of material. That requires an assay that can generate very bright signals to cut through the noise. It’s here that the second breakthrough occurred, as Xiao Wang, working with Karl, developed the STARmap assay. This innovative system for spatial transcriptomics combines dual-probe binding, a ligation step, and signal amplification. STARmap produces extremely bright signals with excellent specificity and sensitivity, enabling 3D imaging of thick tissue at high resolution.
After Xiao moved to The Broad Institute, her team developed an extension of the STARmap assay, called RIBOmap, which was published in Science in late 2023. RIBOmap includes an additional probe that enables the detection of RNAs actively being translated into proteins. This means that instead of just knowing which transcripts are present in a cell, you can now see which ones are functionally relevant – essentially acting as a high-plex proxy for proteomics, but without needing to develop hundreds or thousands of custom antibodies. Both STARmap and RIBOmap can be run on our Pyxa instrument.
We heard from many researchers who loved the data but found the assays too difficult to run in their labs. These assays are extremely labor-intensive and technically complex. That became a key part of our mission at Stellaromics: to take these powerful tools, simplify them, and make them accessible to anyone. To achieve this, we developed a fully automated, highly sophisticated instrument with onboard fluidics, capable of imaging deep into tissue at subcellular resolution.
Our Pyxa platform introduces a fully custom, ultra-high-resolution, massively parallel, high-throughput confocal in situ spatial sequencer to the market. It’s the result of incredible engineering, hard work, and problem-solving by our amazing team, with support from our technology partners.
What are the main challenges in single-cell research today, both in standard single-cell analysis and in 3D spatial omics?
The biggest challenges that we consistently hear from across the single-cell research community are those concerning data complexity and analysis, sample prep, standardization, cost, and multi-omic integration. Standardization remains an ongoing issue. Each lab has a slightly different workflow and employs various methods for measuring and analyzing data. The sheer volume and dimensionality of the data can make it challenging to analyze and interpret. There is still no universal framework, which makes cross-study comparisons difficult.
Sample preparation also remains challenging, although the specifics differ slightly from platform to platform. In single-cell analysis, for example, when extracting cells from complex tissues, the risk of disrupting gene expression is present. In spatial biology, the primary challenges in sample preparation are related to tissue handling, specifically fixation, slicing, sectioning, and processing the material without compromising sample integrity.
The cost of running these assays can be prohibitive for many labs. It’s not just the price of acquiring an instrument but also the per-sample expense of running these assays, especially when scaling up for large studies. Finally, integrating multi-omic data—combining transcriptomics with proteomics, epigenomics, or spatial data—is still a major challenge. It’s the holy grail in many ways, but it’s not yet seamless.
How vital is standardization in the development of novel spatial omics platforms?
Standardization is a critical issue and has been a part of every wave of genomic technology. If you look back at microarrays and the early days of sequencing, we also needed to develop standards in those areas. That’s where groups like Genome in a Bottle, out of NIST, and others came in, working to establish reference materials and protocols across vendors and platforms. It was essential to advance the field.
Similar efforts are starting to emerge in spatial biology, but the challenge is that the technology advances so fast. Standards and metrics become rapidly outdated because a new platform has expanded what’s possible.
Three-dimensional analysis opens the door to defining new, more meaningful metrics, beyond current benchmarks such as transcripts per cell. These technologies allow you to quantify cell-cell interactions in 3D space or assess vertical tissue architecture. These are things you couldn’t measure accurately until now. Standardization is critical, but it’s also a moving target. That has always been true in genomics and will remain so as the field continues to evolve.
How do you anticipate the synergy between single-cell analysis and omics technologies evolving?
These two technologies will likely merge more over time. We’re already seeing signs of this, and I believe the process will only accelerate.
The first step in this fusion is to improve analysis tools and platforms that facilitate the overlay of single-cell data with spatial data. It begins with the ability to compare and correlate these datasets side by side in a meaningful way. Over time, I believe we’ll see the actual technologies start to merge as well.
We will also see more meaningful integration of multiple omics layers, such as epigenomics, proteomics, and metabolomics. Additionally, with our RIBOmap assay, which we refer to as “spatial translatomics,” capturing actively translating RNAs, these key data types will be increasingly included in unified multi-omics frameworks.
And you can’t discuss the future of this field without mentioning AI. AI-driven models will play a significant role in uncovering cellular interactions and patterns that may otherwise remain undetected. As datasets grow in complexity and scale, AI will enable researchers to extract insights from both single-cell and spatial data in entirely new ways. Overall, I believe what we'll see is not just the merging of single-cell and spatial technologies, but also a broader shift toward deeply integrated, AI-enabled multi-omics platforms.
Could you tell me about the potential impact of single-cell analysis, for example, on precision medicine or other areas? What significant problems could it help the field address or resolve?
The primary reason I work in this field, and why many others do as well, is that the potential of spatially-driven single-cell analysis to genuinely transform healthcare is immense. For example, in precision medicine, one of the biggest hurdles is the extreme heterogeneity of diseases. Cancer is a prime example - what works for one patient often doesn’t work for another. To address that, we must be able to identify molecular subtypes and variations within diseases. Spatial mapping and single-cell analysis can aid in this by offering significantly higher specificity and resolution. This, in turn, paves the way for more personalized drugs and treatment strategies.
Take, for example, the tumor.microenvironment. We now understand that the immune, stromal, and vascular cells within and surrounding the tumor are essential for understanding how that tumor behaves and how it might respond to treatment. Spatial biology broadens this view, and studying the tumor microenvironment in 3D will be even more powerful. It enables us to identify relevant drug targets, and design therapies that are more precisely tailored to the patient.
Another significant challenge in the clinical setting is the development of drug resistance. Cancers evolve rapidly and often become resistant to treatments. Using single-cell and spatial analyses allows us to identify subpopulations of resistant cells before they proliferate and metastasize. Using 3D spatial technologies to monitor a larger area of the tumor for these rare, resistant cells could help us prevent treatment failure or relapse. It is also very promising for monitoring minimal residual disease (MRD) after treatment and identifying early signs of recurrence.
Beyond oncology, there are significant implications for autoimmune diseases and the field of neuroscience. In neurodegenerative conditions like Alzheimer’s disease and Parkinson’s disease, understanding tissue heterogeneity and the dynamics of specific cell types is essential. Tracking microglial activation in Alzheimer’s, for example, is a crucial area of research. Spatial biology enables monitoring in ways the field has not previously been able to do. It offers the opportunity to identify therapeutic targets before extensive neuronal loss occurs.
The potential impact is enormous; now, it’s about making the technology more accessible, easier to use, and scalable. Once that happens, we’ll be able to unlock many vital insights in human health.
Looking towards the future, what’s your long-term vision for the field over the next 5 to 10 years?
We can view it from both the technological revolution and application perspectives. From a technology standpoint, as workflows are simplified and data analysis pipelines become more mainstream, I believe spatial will become more accessible and widely adopted. I’m personally excited for our team to help drive that process. Ultimately, I believe spatial and single-cell technologies will become as ubiquitous as next-generation sequencing. You'll start to see one of these instruments in every major lab around the world.
The technology itself will continue to evolve. We're not just talking about moving into three dimensions; we will also see deeper integration of multiple data types, as discussed earlier. There will be more automation, better ease of use, higher throughput, and lower running costs. It will follow a similar trajectory to sequencing and other technologies that have gone through that maturation curve.
You’ll also see different instruments emerge to serve various markets – from ultra-high-throughput spatial sequencers to compact benchtop systems for smaller labs. AI will continue to play a significant role. We're already using AI to build and refine our algorithms. In the future, it will be increasingly used to extract insights from data in a powerful and scalable manner.
From the application side, we’re just beginning, and I expect the range of available applications to grow dramatically over the next decade. Spatial technologies are already making progress in drug development, but I believe they will become essential tools. Understanding how drugs interact with tissues—whether they reach the right targets, how deeply they penetrate, and at what concentrations—is crucial. Spatial methods, especially in 3D, allow for high-resolution payload mapping that was previously impossible.
However, the key question is how quickly spatial technology will become standard in the clinic. Its real promise is in enabling discoveries and medicines by mapping both diseased and healthy tissues with unprecedented detail.
Over the next five to ten years, I expect to see landmark papers that demonstrate the clinical application of spatial data, not only in research but also in diagnostics and patient stratification. We'll see spatial signatures used to identify diseases earlier, guide treatment decisions, and ultimately improve patient outcomes.
Lastly, there is exciting work being done by initiatives like the Human Cell Atlas and other large-scale efforts to map entire human organs at the single-cell and subcellular levels. These projects are poised to redefine our understanding of human biology and disease, and we’re genuinely thrilled to be part of that movement.
Newsletters
Receive the latest analytical science news, personalities, education, and career development – weekly to your inbox.
