Stellaromics emerges from stealth with a bold pitch: forget flat tissue slices, the real action is in three dimensions
The spatial biology wars just got another combatant, and this one's bringing a different playbook.
While competitors spend February's AGBT General Meeting in Orlando touting ever-higher resolution on razor-thin tissue sections, Stellaromics—a Stanford and MIT offspring that's been operating in relative quiet—chose the same week to unveil Pyxa, a platform it claims can peer into tissue samples up to 100 micrometers thick. That's ten to forty times deeper than the industry standard, depending on who you're comparing it to.
It's a provocative angle in a field where most energy has gone toward cramming more genes onto standard pathology slides. The question, of course, is whether anyone really needs that third dimension badly enough to bet on an unproven entrant.
"We're not just capturing more data," CEO Todd Dickinson told analysts in materials accompanying the February 19 launch. "We're capturing different biology." Dickinson, a veteran of Illumina and Bionano, is careful to frame this not as incremental improvement but category creation—always a riskier pitch, but one that commands attention if it lands.
The Technical Gambit
Here's what Stellaromics is actually selling: an integrated system that handles everything from sample prep through analysis, built around proprietary SNAIL probes that use intramolecular ligation chemistry. Researchers feed it tissue sections anywhere from 20 to 100 micrometers thick—versus the 5 to 10 micrometers that works with, say, 10x Genomics' Xenium or Vizgen's MERSCOPE—and the platform uses automated volumetric confocal imaging paired with sequencing-by-ligation to decode spatial gene expression.
The platform accommodates human, mouse, and non-human primate tissue in both fresh-frozen and FFPE formats. It targets short and long RNAs, including siRNA and guide RNAs. Sample prep runs in 12-well microplate format. The company says it achieves subcellular resolution for "hundreds of targets simultaneously," though they haven't put a precise gene count on the record yet.
The promise, according to Stellaromics, is access to biology that simply doesn't exist in two dimensions: immune cells migrating through tumor depth, neurons threading through brain tissue, cell-cell interactions that span vertical space. Serial sectioning and computational stitching—the workaround most competitors use for 3D—introduces artifacts, the company argues, and misses transient structures.
Whether that argument holds up under scrutiny is another matter. Stellaromics hasn't released head-to-head comparison data yet, though peer-reviewed datasets are expected to surface at AGBT, per launch materials.
Early Believers (and Maybe Guinea Pigs)
Before going commercial, Stellaromics placed three instruments in beta programs at research heavyweights. The University of Glasgow took delivery of the first system in October 2025; oncologist Nigel Jamieson is using it to dissect tumor heterogeneity. UC Irvine's Rui Chen, an ophthalmology researcher contributing to the Human Cell Atlas, installed the second in December. Emory University's Hailing Shi—focused on RNA regulation and neuroscience—got the third in late January.
These aren't random picks. All three are studying biological questions where tissue architecture arguably matters as much as molecular inventory: how tumors organize themselves in space, how retinal cells layer and communicate, how neural circuits wire up. If Pyxa can't prove itself useful in those contexts, the value proposition gets shaky fast.
For labs reluctant to commit capital, Stellaromics also offers "Technology Access Services," a sample-to-answer program where the company handles sectioning, imaging, and analysis, then hands back cell-by-gene matrices with XYZ coordinates and subcellular localization. It's a savvy move—pharma companies love to try before they buy, particularly with spatial platforms where sample quality and workflow complexity can make or break adoption.
Walking Into a Firefight
The timing is aggressive, perhaps more than Dickinson expected when he mapped out this launch. Spatial transcriptomics has become one of the hottest segments in life science tools, and February 2026 saw nearly every major player stake fresh claims.
10x Genomics still owns the largest installed base with Xenium, which handles up to 5,000 genes on standard sections and has become the default choice for labs wanting panel breadth without too much operational pain. Vizgen's MERSCOPE uses MERFISH imaging chemistry to cover large tissue areas on flat sections. Bruker layers in proteomics with CosMx and GeoMx. Illumina announced plans in late 2025 to bring whole-transcriptome spatial capabilities to its sequencing platforms this year, with "3D reconstructions" achieved through computational stitching of serial sections—exactly the approach Stellaromics wants to position against.
Then there's Singular Genomics, marketing its G4X system as an in situ multiomics workhorse. And that's just the commercial side. Academic labs are churning out new spatial methods faster than anyone can track.
Stellaromics is betting that native 3D capture represents a fundamentally different value—not just better, but different. That's a harder sell than "we're 10x faster" or "we're half the price," because it requires convincing buyers they have problems they don't know they have yet.
The Money Behind It
The company isn't exactly starving. Stellaromics closed an $80 million Series B last February, led by Catalyst4 with participation from Stanford University Ventures. That came on top of a $25 million Series A in November 2023. For a company just now going commercial, that's significant firepower—enough to fund manufacturing scale-up, field a sales team, and weather the inevitable early-stage bumps.
Dickinson says they're "now shipping" and have opened a limited Early Access Program, though order volume and backlog remain under wraps. The company is actively hiring for commercial roles: an Area Sales Manager for the U.S. Central region, a VP or SVP of Commercial Operations in Boston. Job postings reference capital equipment sales cycles and service contracts, suggesting a hybrid revenue model that blends instrument sales with consumables and services.
What they're not saying: pricing, reagent costs per sample, or delivery timelines. That silence is conspicuous in a market where price transparency increasingly matters, especially as spatial platforms compete with cheaper RNA-seq alternatives for budget dollars.
What We Still Don't Know
Launch materials don't include independent benchmarking—no sensitivity comparisons, error rates, or cost-per-sample data against Xenium or MERSCOPE. Regulatory status isn't specified; Pyxa appears research-use-only for now, which is standard but limits near-term clinical applications. And while three beta sites is a decent start, it's hardly a wave of adoption.
There's also the workflow question. Spatial platforms notoriously demand finicky sample handling and deep technical expertise. If Pyxa adds complexity atop that—thicker sections can be harder to work with—adoption could stall regardless of data quality.
Still, for certain research questions, Stellaromics may have identified a genuine gap. Organoid biology. Brain circuit mapping. Tumor microenvironment studies where three-dimensional architecture isn't a nice-to-have but the whole point. The labs studying those systems have been making do with serial sectioning or clearing methods that destroy molecular information. A native 3D platform, if it works as advertised, could be legitimately enabling.
The harder question is whether that niche is big enough to sustain a business against entrenched competitors with broader applicability. Perhaps in a few quarters we'll know whether "going deep" was a clever market entry or a bet that missed the mark. For now, Stellaromics has put its chips on the table.
