Last week, during a meeting, I heard some researchers say:

“We can already do spatial transcriptomics without expensive technology or lab experiments.”

My immediate thought: I wish we could.

When I asked how, the answer was simply: “With AI.”

That wasn’t exactly convincing for me. 😅 So I went digging.

And I found some cool deep-learning approaches that might explain where this idea comes from. But there’s an important distinction:

Predicting spatial gene expression is not the same as performing a spatial transcriptomics experiment.

And there’s another important point: these AI models need to be trained on data generated by… you guessed it…“expensive” spatial transcriptomics wet-lab experiments.

Could that change in the future?

Maybe.

As AI models become better trained across different tissues, diseases, and biological contexts, perhaps we will need to produce fewer experimental spatial transcriptomics datasets.

My P.O.V.: That’s where we’re heading.

Anyways, today’s video is about Path2Space, a deep-learning model that predicts spatial gene expression directly from standard H&E histopathology slides.

🎥 After today’s video you will know:

• Why Path2Space was developed

• How it works

• What makes this different from spatial transcriptomics lab experiments

• Why approaches like this could be important for cancer research & precision oncology

Watch it now!

If you liked it, just hit ❤️.

And don’t forget to tell me what topic you’d like next. 👇🏻

Start Survey

Thanks!

Cátia

Let’s connect on LinkedIn

Hi everyone,

The past few weeks have been incredibly busy. Between interviews, a new job offer, and preparing for a big move, I haven’t had much time to write ✍️ or keep up with everything happening in the single-cell and spatial biology world.

The biggest news is that, in September, I’ll be joining IRB Barcelona as Head of the Advanced Genomic Technologies Core Facility. It’s an exciting new chapter, and 1 I’m really looking forward to.

August will be a transition month. I’ll be taking a short break, looking for an apartment, settling into Barcelona, and thinking about the future of The Single Cell World.

Many of you have asked if I’ll continue with it.

The answer is yes.

What I don’t know yet is what the format will look. My new role will naturally change how I spend my time, and I want to make sure everything I publish remains useful, independent, and worth your time.

Thank you for reading, sharing, commenting, and supporting this project over the past years. It truly means a lot!

Now, let’s catch up on everything that happened in the single-cell and spatial biology world this month.

✍️ This Month’s Deep Dives

🔍 Other Things Worth Of Knowing

1. Human Cell Atlas General Meeting 2026

Human Cell Atlas 2.0 is becoming spatial + clinical.

2. Project JAGUAR

👀 Only ~2% of genomic datasets come from Latin America.

Project JAGUAR aims to change that with a large-scale single-cell immune atlas

💬 Comments From Our Community (1950 subscribers)

“The single-cell items are always the highlight, but that SpatialBench-Long number is the one I keep turning over. 11% percent isn’t really about models being weak at finding signal in spatial data. The wall is the step after: deciding which signal is the actual discovery and which is an artifact of the tissue or the segmentation.

That judgment is what separates analysis from discovery, and it doesn’t fall out of the data no matter how much you have.”

Dr Efevretis

Leave a comment

📅 Planned For Next Month

• 🌴 A short summer break.

• 🏡 Finding an apartment in Barcelona.

• 💡 Planning the next year for The Single Cell World.

See you on September. ✨

ChronoSeq introduces a new approach to single-cell transcriptomics by adding the missing dimension: time.

🎥 In today’s video, I explain this 👇 bioRxiv paper in simple language. 😉

You will know
🔹How ChronoSeq works
🔹The biological rationale
🔹 The potential applications

Watch it now!

If you liked it, just hit ❤️.

And don’t forget to tell me what topic you’d like next. 👇🏻

Start Survey

Thanks!

Cátia

Let’s connect on LinkedIn.

🔍 Other Things Worth Of Knowing

1. Parse Biosciences & bit.bio

Parse Biosciences (now part of QIAGEN) and bit.bio announced a collaboration to build a map of transcription factor-driven cell identity.

Bit.bio is a synthetic biology and biotechnology company. They sell pre-programmed human cells (called ioCells).

The goal?

➡️ Understand how specific transcription factors control cell fate.

➡️ Test thousands of genetic perturbations at scale.

➡️ Create datasets that link genetic changes to cellular outcomes.

Why does this matter?

👉 Many companies are talking about AI in biology.

But AI is only as good as the data used to train it.

Building large-scale maps that connect genetic inputs to cellular outputs may become one of the most valuable resources in the field.

2. 🤖 Can AI act as an independent scientist?

Not yet. In a new benchmark called SpatialBench-Long, the best models recovered scientific discoveries from raw spatial biological data only 11.1% of the time.

3. SeekGene New Products

👀 Every 6 months I check what single-cell companies are building.

This time, I was surprised by SeekGene.

Their portfolio now spans sample prep, single-cell, spatial biology, and multiomics. Also data analysis.

Interesting to see companies evolving into end-to-end platforms.

4. Spatial Transcriptomics–Guided Docetaxel Therapy Stratification in Prostate Cancer (𝗦𝗧-𝗗𝗼𝘅𝗣𝗖𝗮)

ST-DoxPCa uses spatial transcriptomics-guided pathology to predict docetaxel benefit in metastatic hormone-sensitive prostate cancer.

AI model + spatial transcriptomics + pathology + chemotherapy

5. Haga Bio Company

Interesting company to keep an eye on.

🎯 Their Goal: Validate candidate biomarkers directly in real tissue

• In situ RNA detection

• Single-nucleotide precision

• Single-cell resolution

• Works with FFPE and archived tissue samples

• No specialized instrument, only existing histology and pathology lab infrastructure

Interesting company to keep an eye on.

6. Novogene Restrictions at U.S.

The U.S. Department of Defense (DoD) added Novogene to its Section 1260H list of "Chinese Military Companies.” This designation links the company to emerging federal procurement restrictions, signaling increasing national security scrutiny over U.S. and Chinese life sciences collaboration

7. 10x acquires Proteintech Genomics

More protein + RNA measurements point toward a more multimodal/multiomics future in single-cell and spatial biology.

💬 Comments From Our Community (1904 subscribers)

“Another important feature, besides transcript detection sensitivity in spatial biology, is the accuracy of cell segmentation. If individual cells cannot be precisely identified, the number of detected transcripts becomes less meaningful.”

“Who thinks that 18,000 genes represent the whole transcriptome?”

📅 Planned For Next Month

• ChronoSeq

• StrataMap vs. competitors

• Why Do Whole Transcriptome Kits Still Show 3′ Bias?

• July 2026 Newsletter

And much more…

You can also tell me what you would like to read or see next. 👇

Start Survey

If you’ve been finding this useful, consider supporting the work as a paid subscriber. 💗

See you next month. ✨

For months, the company had been teasing its entry into the spatial transcriptomics market. Presentations appeared at conferences. Early-access users started sharing preliminary results. Rumors circulated about a new platform built around Illumina’s sequencing ecosystem.

June 8, 2026, Illumina finally revealed it.

StrataMap Spatial Solution.

Today, I want to summarize everything I know so far, what makes it different, and some of the questions that for me still remain unanswered.

Meet StrataMap

StrataMap Spatial Solution is not an instrument. It is an ex situ1 sequencing-based, end-to-end whole-transcriptome spatial transcriptomics workflow designed to run on existing Illumina sequencing platforms.

Key features:

• Whole-transcriptome spatial profiling

• Poly(A)-based RNA capture

• Sequencing-based approach

• Single-cell resolution

• High‑density 1 µm spatial features

• Coding & long noncoding RNA with poly(A). Can also detect noncoding genes and pseudogenes

• Species agnostic for polyadenylated transcripts. Illumina has already demonstrated the technology in human, mouse, pig, rhesus macaque, marmoset, fruit fly, corn, and rhinoceros samples

• Fresh frozen samples; compatible with optimal cutting temperature (OCT) tissue microarrays (TMA)

• Compatible with enrichment workflows such as those used for VDJ clonotypes

• Market larger imaging capture area (50 × 15 mm; 750 mm² per slide)

  

• Integration with DRAGEN plus Illumina Connected Multiomics analysis

Future Roadmap:

• Formalin-Fixed Paraffin-Embedded (FFPE) samples coming in 2027

How Does StrataMap Work?

1. Tissue preparation and imaging

Fresh-frozen tissue sections are mounted onto a spatially barcoded capture surface slide, stained, and imaged to preserve tissue morphology and establish the spatial reference.

2. RNA capture and library preparation

The tissue is permeabilized, allowing poly(A) transcripts to bind to spatially barcoded oligonucleotides distributed across continuous 1 µm surface features. Captured RNA is converted into spatially indexed cDNA and prepared as sequencing libraries.

3. Sequencing on Illumina instruments

Libraries are sequenced on compatible Illumina platforms, including the NextSeq 2000, NovaSeq 6000, NovaSeq X, and NovaSeq X Plus. Sequencing depth can be adjusted according to tissue complexity and study design.

4. Data processing

Sequencing data are processed using the DRAGEN spatial transcriptomics pipeline to generate spatial gene expression maps. Results can then be explored in Illumina Connected Multiomics for clustering, cell-type identification, and data visualization.

Some other details:

• Assay time: 2.5 days

• Hands-on time: ~8.5 hours

• Input Quantity: 1-12 sections

• Multiple slides can be run in parallel

What will you need?

Looking at StrataMap as a Researcher

Read more

“Another cell atlas?”

Haven’t we already built several of those? Big single-cell datasets?

• The Human Cell Atlas.

• Different organism-specific atlases.

• The Chan Zuckerberg Billion Cell Project.

• The 100 Million Cell Challenge project from Scale Biosciences.

And now the Illumina’s Billion Cell Atlas?

At first, to be honest, they can sound similar. After all, they all involve massive datasets, millions or even billions of cells, and ambitious efforts to better understand biology.

But after reading about each 1 of these initiatives, I realized they are not trying to accomplish the same. In fact, each 1 reflects a different stage in the evolution of single-cell biology and a different vision for its future.

Understanding those differences tells us a lot about where the field is heading next.

And that’s exactly what I want to talk about today.

1. Mapping the Cells of the Human Body

In 2016, the Human Cell Atlas was launched with an ambitious goal:

Create a reference map of all human cell types.

The idea was simple but transformative.

If the Human Genome Project provided a reference sequence of human DNA, the Human Cell Atlas aimed to catalog the cellular diversity of the human body.

Researchers from around the world—including myself 😉—began generating single-cell datasets across tissues, organs, developmental stages, and diseases.

The main question was:

What cells exist in the human body?

For much of the last decade, this question drove the field forward.

The Human Cell Atlas also inspired a wave of other organism atlas projects. The goal was largely the same: identify and catalog cellular diversity.

Some examples:

• Mouse Cell Atlas - Tabula Muris

• Fly Cell Atlas

• Zebrafish Cell Atlas

• Plant Cell Atlas

• C. elegans Cell Atlas

2. Bigger Datasets Become Possible

As single-cell technologies matured, the number of cells that could be profiled increased dramatically. What once required months of work could now be achieved in days.

Thousands of cells became tens of thousands. Tens of thousands became millions.

The conversation began to shift.

Researchers were no longer asking whether large studies were possible. They were asking how far throughput could be pushed.

The key question became:

How many cells can we measure?

Projects such as Scale Biosciences’ 100 Million Cell Challenge (2024) demonstrated that profiling tens or even hundreds of millions of cells was becoming technically feasible. Supported by organizations like the Chan Zuckerberg Initiative (CZI), the project highlighted just how rapidly single-cell technologies were advancing.

The goal was not to build a comprehensive atlas but to demonstrate what was now possible at scale.

These efforts naturally led to the next era of projects.

3. Enter the Billion-Cell Era

As datasets continued growing, organizations started thinking even bigger.

Read more

“CosMx was already doing this.”

At first glance, it may sound like frustration over Atera’s release. But behind that comment is actually a valid question:

Is it true?

If Atera promises whole-transcriptome spatial biology at single-cell resolution and scale, then it’s worth asking whether CosMx SMI (CosMx), from Bruker Spatial Biology, was already delivering it, being the impossible already possible.

Now, before diving in, let’s be clear: the real comparison between these 2 platforms, for me, cannot be solved by “online research” alone.

A proper head-to-head evaluation would require serial sections of the same tissue (s), processed side by side, by the same operators, under identical conditions, followed by blinded data analysis.

That is the only way to objectively compare performance.

As many of you know, I’m no longer at the bench, so I can’t generate that dataset myself. But I’m sure some of you eventually will.

What I can do is something simpler:

Take each core promise behind Atera and ask 1 question:

Was CosMx Already Doing This?

This blog is not about which platform is “better.” It’s about understanding:

1. where they overlap.

2. where they differ.

3. whether Atera is introducing something genuinely new or packaging existing capabilities in a different way.

Let’s get into it!

Atera Promise 1: Whole-transcriptome spatial biology at single-cell resolution

Can CosMx already do this?

Read more

🚀 Meetings, launches, papers.

Get your cup of coffee or tea, and let’s get into it.

✍️ This Month’s Deep Dives

Spatial Touchstone: Can We Finally Standardize QC in Spatial Transcriptomics?

🔍 Other Things Worth Of Knowing

⚖️ 10x Genomics and Curio Bioscience are close to settling their patent lawsuit.

• 10x Genomics sued Curio in 2023, alleging the infringement of different spatial biology patents exclusively licensed from Prognosys Biosciences.

• The planned jury trial was canceled.

• Terms are still undisclosed.

• Curio is now owned by Takara Bio.

On the same day, 10x filed another patent lawsuit. This time against Element Biosciences.

The Chan Zuckerberg Biohub is a collaborative research initiative that brings together leading institutions to develop new technologies and generate large-scale biological data to accelerate discoveries in biomedicine and AI-driven biology.

Update about Lightcast Bio. #singlecell #technology 👇

💬 Comments From Our Community (1827 subscribers)

📅 Planned For Next Month

• Atera vs. CosMx

• The Billion Cell Atlas:

• 3′ vs Whole Transcriptome

• ChronoSeq

• June 2026 Newsletter

And much more…

You can also tell me what you would like to read or see next. 👇

Start Survey

If you’ve been finding this useful, consider supporting the work as a paid subscriber. 💗

See you next month. ✨

Led by Dr. Jasmine Plummer, Director of Center for Spatial Omics at St. Jude Children’s Research Hospital, this collaborative initiative aims to standardize imaging-based spatial transcriptomics through reproducible workflows, quality metrics, and cross-platform comparisons.

🎥 In today’s video, I break down in simple language 😉 this Nature Biotechnology paper. 👇

I tell you why it matters and how it could improve the way we assess quality, reproducibility, and confidence in imaging-based spatial transcriptomics data.

Watch it now!

Read more

While going through summaries from recent life sciences meetings, I noticed how often certain details were highlighted.

Not only technologies or datasets.
But collaborations. 🤝

📰 For example:

“10x Genomics, Inc. (Nasdaq: TXG), a leader in single cell and spatial biology, today announced a collaboration with Brigham & Women's Hospital, a world-class academic medical center….” or “announced a collaboration with Dana-Farber Cancer…”

Source

And elsewhere:

“The Illumina Billion Cell Atlas is the first tranche of its program to build a 5 billion cell atlas over three years […] Under an alliance framework with AstraZeneca, Merck, and Eli Lilly and Company leading as founding participants…”

Source

These are, on the surface, perfectly standard announcements.
Factual. Professional. Expected.

But reading several of them back-to-back, something subtle stood out.

What is Actually Being Highlighted?

In many of these updates, the emphasis isn’t only on the technology.

It’s also about who is involved. 👥

• Major medical centers 🏥

• Well-known research hospitals 🧬

• Large academic initiatives 🎓

When these names appear next to a platform or a dataset, something subtle but powerful happens.

• The technology gains legitimacy since it has been trusted by groups with high standards

• The workflow feels safer

• The adoption risk feels lower

No one is explicitly telling others what to adopt.

No one is promoting anything.

And yet, a signal is sent.

Where Researchers Come In (often unintentionally)?

Most researchers involved in these collaborations would probably say:

“I’m not an influencer.”

And they’re right, in the usual sense of the word.

• They are not trying to persuade

• They are not endorsing products publicly

• They are simply doing good science, often through partnerships that make that science possible.

Still, those decisions matter.

Agreeing to collaborate, generate data, or associate with a platform shapes how the wider community perceives that technology. 👀

• Other labs notice

• Reviewers notice

• Companies notice

This influence isn’t intentional. It’s not even visible most of the time.

It’s a byproduct of trust and visibility in a growing ecosystem.

The Quiet Impact

Researchers just… agreed to collaborate. And yet, that agreement alone influences:

• What other labs adopt

• What reviewers consider acceptable

• What companies prioritize next

That’s when it clicked.

In single-cell and spatial transcriptomics, influence doesn’t come from visibility.
It comes from association.

And researchers, whether they like the label or not, are central to that process.

A (Gentle) Conclusion

This isn’t about blame, labels, or turning researchers into marketers.
It’s about awareness.

Single-cell and spatial transcriptomics have reached a point where choices made quietly, through agreements and collaborations, are shaping the direction of the field.

And maybe the most interesting part is this:

Most researchers may not see themselves as influencers. But unfortunately, influence doesn’t require intention.

I hope you like it. If you did, just hit ❤️ & share it.

Thanks for reading it.

And don’t forget to tell me what topic you’d like next. 👇🏻

Start Survey

Cátia

Let’s connect on LinkedIn.

Back then, I was working in cancer & epigenetics.

Single-cell? Spatial transcriptomics?
Not really a thing yet…

Honestly, I don’t think anyone imagined what was about to come.

Fast forward to AACR 2026.

If you followed it, it probably felt like everything revolved around 1 announcement:

• Atera from 10x Genomics

And yes, it was a big moment.

But no, the AACR 2026 was not just about that.

A lot of other important things happened (Figure 1).

Let’s take a look.

1. New Commercial Options

Several companies introduced new tools expanding what’s possible in single-cell & spatial biology.

QIAGEN (Parse Biosciences)

Evercode WT v4

• Improved sensitivity + faster workflow

Evercode WT FFPE kit

• Scalable instrument-free single-cell + FFPE compatibility

  (Abstract 4937: multiple organs)

Read more

🚀 New launches. Big promises. Subtle shifts.

Get your cup of coffee or tea, and let’s get into it.

✍️ This Month’s Deep Dives

SPONSORED BLOG by “Flossing With Cajal” publisher.

🔍 Other Things Worth Of Knowing

1. AtlasXomics

Spatial epigenomics is possible now.

2. Stellaromics

3D spatial biology → slowly becoming real.

• Moving beyond 2D slices

• Capturing full tissue structure

👉 Worth watching.

3. Single-cell from FFPE is no longer a 1-player space

Parse Biosciences expanded Evercode WT to FFPE.

Why does it matter?

Until now → 10x Genomics dominated.
Now → there’s a choice.

🛠️ Tips & Resources

1. Using an automatic cell counter for your single-cell experiment?

If you’re switching between sample types and using an automatic cell counter, calibrate it. Seriously.

⚠️ A counter optimized for small cells (e.g., PBMCs) will “lie” to you when you throw big cells (e.g., hepatocytes) at it.

⏱️ A few minutes of calibration can save you from a ruined single-cell experiment.

2. Sample Reality Check

What single-cell companies will ask you: >95% viability, 0% debris, and suspension of individual cells.

What you should aim for: >70% viability, <10% debris, and suspension of individual cells.

3. 🌱 Plant Single-Cell Resources

2 time-savers:

• PlantCellMarker

• PlantscRNAdb

💬 Comments From Our Community (1766 subscribers)

📅 Planned For Next Month

• AACR 2026

• Spatial Transcriptomics Standardization Tools

• The Hidden Influencers in Single-Cell & Spatial Transcriptomics

• May 2026 Newsletter

And much more…

You can also tell me what you would like to read or see next. 👇

Start Survey

If you’ve been finding this useful, consider supporting the work as a paid subscriber 💗

Subscribe now

See you next month. ✨

🍫Chocolate bars.

📅 A single date: April 18, 2026.

👀 And 1 big question:

What would 10x Genomics actually release?

Now we know. Atera.

But the real question is: Were Atera’s “promises” impossible before?

Meet Atera

Atera is the newest spatial instrument from 10x Genomics (10x), joining Visium HD and Xenium.

Key features:

• Whole-transcriptome (WT) (~18,000 genes) with optional stacking of 1000-gene panels

• Compatible with human samples1

• Fresh frozen and FFPE samples, Tissue Microarrays (TMA)

• Tissues and cell lines

• Single-cell resolution

• Available gene panels: Custom, Human Multi-Tissue, and Cell Typing2

• Large imaging areas (>500 mm² per slide)

• High throughput (4 slides/run), 7 days/run

• Standard, off-the-shelf slide compatibility

Future Roadmap:

• Automation

• WT for other species (e.g., mouse WT)

• New gene panels (e.g., human brain/neuro, human immuno-oncology)

• Base-to-base sequencing for mutation detections

• Protein multiomics

2

For the moment.

---

The Hype

If you watched the Atera marketing campaign, it’s easy to feel like…

“This changes everything.”

“The impossible was made possible.”

But let’s slow down for a second.

Atera is an amazing instrument, no doubt. But most of its promises were already possible individually:

• WT at single-cell resolution → already possible (e.g., Visium HD, STOmics)

• True single-cell / high-resolution spatial → already possible (e.g., Xenium, CosMx, MERSCOPE)

• Imaging-based assays → already possible (e.g., Xenium, CosMx, MERSCOPE)

Also, to better understand where Atera is positioned right now compared to 10x existing spatial transcriptomics platforms, Visium HD and Xenium, I prepared Table 1.

---

So, What Actually Changes?

Despite the overlap in features between Visium HD, Xenium, and Atera, this last one will allow us to (Figure 2):

• Scale up spatial transcriptomics projects in a less time-consuming and maybe less expensive way

• Higher resolution

• Enable more systematic, cohort-level analyses

And now I leave you with a quote from Dr. Jan-Philipp Mallm, head of the DKFZ Single-cell Open Lab, German Cancer Research Center, and 1 of the first users.

“Atera enables us to generate data across larger cohorts at a pace and depth that was previously not practical […]”

The ability to measure whole-transcriptome data at single-cell resolution directly in intact tissue represents a long-standing goal for the field and has the potential to significantly accelerate discovery.”

Dr. Jan-Philipp Mallm,

Head of the DKFZ Single-cell Open Lab, DKFZ.

---

The Strategic Reality

Scale is powerful. But this is where caution is needed.

To really benefit from Atera, in my opinion, you need:

• A high number of samples

• A cohort-based biological question

• Infrastructure to handle large datasets

• Budget (instrument, consumables, analysis)

If you don’t have this?

👉 You are probably not the target buyer. A user?

If your project genuinely requires this type of experiment, there are alternative routes.

For example, 10x already offers Atera through its Catalyst Research Services, where samples can be submitted and processed externally. Over time, I assume that additional service providers will offer the same.

---

If I Needed to Invest My Money Today

Case 1: No single-cell or spatial capabilities in the lab + available budget

I would not invest in Atera over the next 2–3 years.

Instead, I would invest in:

• 1 single-cell RNA-seq platform or assay

• 1 spatial transcriptomics platform for discovery (WT)

• 1 spatial transcriptomics platform for validation (in situ)

Why?

• Proven performance

• Established workflows

• Lower scientific risk

• Strong community adoption

This combination allows a lab to:

• Build technical expertise

• Troubleshoot with existing support networks

• Establish collaborations

• Generate a solid track record in single-cell and spatial biology

Case 2: Established lab in the field of single cell and spatial transcriptomics + strong budget

Even in this case, I would not invest in Atera yet. Not because it isn’t promising, but because it’s still early.

Atera is arguably right now the most interesting spatial platform for the future.
In the long term, it could evolve into a 3-in-1 system: a single-cell RNA-seq assay + spatial transcriptomics discovery assay + spatial transcriptomics validation assay. With high resolution and increased throughput.

But right now, it still needs to:

• Be broadly shipped and installed

• Be tested across different lab environments

• Generate independent datasets

Realistically, wider validation will only emerge over the next 1-2 years.

And with that will come:

• Unexpected technical challenges

• Workflow refinements

• Protocol improvements

It’s also important to note that Atera is currently limited to human samples and doesn’t yet support protein detection alongside RNA.

While these features are part of 10x planned developments, this is another reason for me to wait.

---

Final Thoughts

Atera makes it feasible to generate whole-transcriptome high-resolution data across large cohorts at a scale that was previously impossible.

This will accelerate efforts such as human atlases and population-scale clinical studies.

Looking ahead, Atera feels like a combination of what Visium HD and Xenium currently offer. If that trajectory holds, these platforms could eventually be discontinued in the long run.

This also raises an important concern.

Access.

Spatial transcriptomics already has a gap between labs that can adopt these technologies and those that cannot. With instruments like Atera, more complex and more expensive, that gap may increase even further. And that has consequences.

It affects:

• Who can generate data

• Who can lead large-scale studies

• Who gets to publish at the highest level

As spatial biology moves toward population-scale studies, the question is no longer only what is possible.

It’s who gets to do it.

Just before I finish. I’m curious to know 👇….

Thanks to our sponsor: the publisher of “Flossing with Cajal .”

“What happens when knowledge itself starts to disappear?

Flossing with Cajal explores a world where science, memory, and discovery become unstable.

If Atera made you think about how we generate knowledge…

👉 This book explores what happens when that knowledge starts disappearing from papers and publications.

Read the first chapters here!

As always, I hope you like it. If you did, just hit ❤️ & share it.

Thanks for being there.

Share

Cátia

Let’s connect on LinkedIn.

Sources: 10x Genomics website, press releases, and my own knowledge.

If you’re interested in sponsoring The Single-Cell World (blog, video, or interview), feel free to reach out (catiamoutinhos@gmail.com).

They sent chocolate bars 🍫 to some attendees’ hotel rooms. Stamped with a single date:

04.18.26

At the same time, their website quietly changed:

“Exciting news coming from us on 04.18.26.”

And the marketing campaign?

“IMPOSSIBLE.” → “IMPOSSIBLE.”

Simple and clever, I have to say. A teaser we couldn’t ignore (Figure 1).

Since then, like any curious researcher, I’ve been trying to gather information. But honestly?

There are no real leaks. Just rumors, gossips, opinions, and non-official signals.

So this is what I’ll share here today.
👉 What I’ve heard
👉 What I think
👉 And why…

Before reading, keep in mind that this blog is based on rumors, opinions, and informal information, so some of it may not be true.

🔍 The Rumors

As I mentioned, there’s a lot of speculation around what will be announced. Here are the main ones and how I see them.

1. Visium HD Expansion

Rumor: Slides with a larger capture area, higher throughput, and improved workflows.

Do I think it will be this? → No

Because this was already made available. Or at least it was presented in 1 of the last conferences I attended (Figure 2).

Right now we can perform Visium HD experiments using:

• 11 × 11 mm capture areas (Visium HD XL)

• Updated protocol (Visium HD WT Panel sensitivity with the new protocol 2.0)

  

2. Improvements or a New Xenium Platform?

Rumor: New Instrument.

• Work with cells, not just tissue sections.

• Whole transcriptome (WT), not only target-based.

• Higher resolution.

• Higher number of detected genes.

• Higher throughput.

Do I think it will be this? → Maybe. But will this be IMPOSSIBLE for me?

Only if I can have all these options in 1 instrument. If not, not really. Let me explain to you why.

1. Work with cells, not just tissue sections.

   1. This can already be done using the STAMP protocol upstream of the Xenium but also from other spatial transcriptomics imaging-based platforms.

2. Whole transcriptome (WT), not only target-based.

   1. This will be amazing. But where do the CytAssist and Visium HD stay? If I can do WT and target-based experiments in 1 instrument, why should I buy them? Will 10x sacrifice it?

3. Higher resolution, higher detected gene plex, and higher throughput.

   1. This for me will be an update of the Xenium. Not something IMPOSSIBLE.

Besides, the Xenium is still recent. Launching an updated system now would be… a difficult message for users who just invested in it. Well, for me it would have been, unless I had the option of “upgrading it” at a reduced cost. Because new technology usually means spending more money.

3. Diagnostic-grade Kits for Clinical Use

Rumor: 10x could release diagnostic-grade kits designed for clinical applications.

Do I think it will be this? → Interesting… but unlikely for now.

There are actually some signals pointing in this direction. For example, at AGBT 2026, 10x announced collaborations with Brigham and Women’s Hospital, and Dana-Farber Cancer Institute.

They also mentioned plans to establish a CLIA-certified laboratory, which would enable the development of clinical diagnostic tests.

This would be a big step.

• Moving from research-use-only to clinical-grade products

• Entering regulated environments

• Competing in a completely different space

And it would fit the “Impossible → Possible” narrative? Yes.

But…

• This is not just a product launch

• It will require regulatory approval, validation, and infrastructure

This feels more like a long-term strategic direction than an April reveal.

🔮 Another Thing(s) I See As IMPOSSIBLE.

1. New Single-cell & Spatial Multiomics Kits

For example, to measure DNA + RNA + protein from the same individual cell? Ideally, DNA methylation? A kind of adaptation for the DOGMA-seq?

• This is still 1 of the hardest problems in the field. Because:

  👉 DNA, RNA, and proteins are difficult to preserve together without compromising data quality.

2. New Single-cell Ultra-throughput Kits

Kits that are a fusion of 10x and Scale Biosciences technologies?

3. New Spatial Platform

Maybe to do spatial 3D experiments?

Well, I see this a little bit difficult also, since there are already companies working in this space, and I expect that legal problems are not desired.

👀 Final Take

• Visium HD improvement → already in place.

• Xenium improvements → likely.

• New spatial platform → only if it changes the game.

• Diagnostic-grade kits for clinical use → interesting, but unlikely.

• Single-cell multiomics → the underrated possibility.

And honestly?

The most interesting scenario is the 1 people are not talking about yet.

📅 What Happens Next

• April 18 → Announcement

• April 21+ → Webinars with details.

That’s when we move from speculation… to reality.

🚀 Next week, I’ll break down what they actually release and what it really means.

Now, before I finish, what do you think 10x will announce?

Thanks for reading it. I hope you liked it.

If you did, just hit ❤️ & share it.

Share

Cátia

Let’s connect on LinkedIn.

Subscribe now

“Wait… wasn’t this company independent?”
“Who owns this technology now?”
“Didn’t they used to be competitors?”

I’ve had it too. It can be… confusing.

Because it’s not just about new technologies being launched. It’s also about companies quietly buying each other.

Over the past few years, a wave of acquisitions has reshaped the field. And unless you’re actively tracking it, it’s easy to lose control of who bought who.

So here’s a simple breakdown. ⬇️

💼 Acquisitions:

• Illumina → acquired Fluent BioSciences

• 10x Genomics → acquired Scale Biosciences

• QIAGEN → acquired Parse Biosciences

• Waters Corporation → acquired BD Biosciences and Diagnostic Solutions

• Takara Bio → acquired Curio Bioscience

• Bruker → acquired NanoString Technologies

• Quanterix → acquired Akoya Biosciences

• Standard BioTools → acquired DVS Sciences

• Bio-Techne → acquired Lunaphore, Namocell, and Advanced Cell Diagnostics

At first glance, these look like standard business moves. However, for me they signal something deeper. And that matters more than most people think.

From Tools to Ecosystems

Single-cell and spatial technologies are no longer just about isolated tools.

Companies are building end-to-end ecosystems:

• Sample preparation

• Library preparation

• Sequencing or imaging

• Data analysis

• Interpretation

Instead of us, researchers, stitching workflows together, companies are trying to own and offer the full pipeline.

That’s why large players are acquiring smaller, specialized innovators.

⚙️ Speed vs. Specialization

Startups like Parse Biosciences, Scale Biosciences, or Curio Bioscience were built around very specific innovations:

• Combinatorial indexing

• Scalable single-cell workflows

• Spatial mapping without complex instrumentation

By acquiring them, larger companies gain:

1. Faster innovation cycles

2. Access to new customer segments

3. Technologies they didn’t build internally

For us, researchers, this can mean more integrated solutions…but also fewer independent options.

Final Thoughts

This consolidation suggests that single-cell and spatial biology are entering a new phase:

• From innovation → standardization

• From fragmented tools → platforms

• From early adoption → broader deployment

In other words, the field is maturing. And with that comes both opportunity... and trade-offs.

In single-cell and spatial biology, the real competition is no longer between technologies. It’s between ecosystems.

I hope you like it. If you did, just hit ❤️ & share it.

Thanks for reading it.

And don’t forget to tell me what topic you’d like next. 👇🏻

Start Survey

But it still comes with limitations:

• cost

• scalability

• the fact that cells are destroyed during the process.

What if you could profile millions of cells… without sequencing?

Read more

My first thought was, "Wait… isn’t this what I was already doing 10 years ago?”

At the time, I was a master’s student in a histopathology department doing cancer research. My days were filled with:

• H&E staining

• Fresh vs. paraffin sections

• FISH and immunofluorescence

• Sitting with a pathologist to choose the right tumor regions

So when spatial transcriptomics started gaining traction, it didn’t feel entirely new. It felt… familiar.

Spatial transcriptomics today:

• Connects gene expression to tissue structure

• Enables in situ biology at scale

But what’s less discussed is:

👉 All the complexity of tissue workflows.

If you haven’t worked with histology before, the learning curve can feel… steeper than expected.

Here are some of the most common problems I see when researchers ask me for help. 👇🏻

🔬 1. You will need pathology and histology expertise

This is not just an RNA experiment. It’s about the tissue.

• Is this the right tissue?

• Are they the right cells?

• Is morphology preserved?

These are pathology questions you need to answer.

And honestly, this is where projects quietly fail:

• Wrong tissue region

• Poor section quality

• Misinterpreted tissue

If no one on your team can read a slide, you’re guessing.

⚡ 2. You must choose: sequencing vs. imaging

There are 2 main spatial transcriptomics approaches:

1. Sequencing-based

• Whole transcriptome

• Lower resolution

• Good for discovery

2. Imaging-based

• High resolution

• Limited gene panels

• Good for validation

So the real question is:

Do you already know what you’re looking for?

🧪 3. RNA quality will make or break everything

RNA in tissue is fragile. Fixation, storage, sectioning…all of it matters.

And spatial assays don’t forgive poor quality.

What helps (and is often skipped):

• Pilot runs

• RNA quality checks (DV200, RIN)

• Protocol testing

🌈 4. Autofluorescence can ruin your readout

Some tissues naturally glow (e.g., brain, lung, FFPE). That background competes with your signal.

Result:

• Lower sensitivity

• More noise

• Harder interpretation

It’s not a dealbreaker, but you need to plan for it early.

🧩 5. Segmentation is still a bottleneck

Even now, 1 thing remains difficult. Defining where one cell ends and another begins.

• Cells overlap

• Boundaries are unclear

• Nucleus ≠ cell

So even with beautiful data, interpretation can get messy.

Make sure you have a clear segmentation strategy, or someone who does.

Final Thoughts

If you want your experiment to work, it’s not just about the technology. It’s about the team.

• Pathologist

• Histology expert

• Bioinformatician

Because every step can fail.

👉 Spatial transcriptomics is not plug-and-play. It’s a multidisciplinary workflow. And the researchers that understand this early get the best results.

I hope you like it. If you did, just hit ❤️ & share it.

Thanks for reading it.

And don’t forget to tell me what topic you’d like next. 👇🏻

Start Survey

Cátia

Let’s connect on LinkedIn.

Subscribe now

Many announcements were delivered as integrated platforms combining instruments, chemistry, and software into a single ecosystem. Several companies also emphasized translation-adjacent features, such as formalin-fixed paraffin-embedded (FFPE) tissue compatibility, workflow standardization, and scaling across large cohorts.

This suggests that spatial biology is moving from small-scale experiments toward production-scale datasets designed for translational research and AI-driven discovery.

So what did AGBT 2026 reveal about where the field is heading?

Let’s take a closer look.

From 2D Sections to 3D and Integrated Ecosystems

Read more

Not because every major launch happens there, but because the meeting usually reveals what the next 6–12 months will look like for the field. It also gives me plenty of topics to write ✍️ about.

This year, the signals around single-cell technologies were subtle but meaningful. Instead of dramatic launches, the focus shifted to:

• 🔬 scaling experiments

• ⚙️ improving workflows

• 🧠 enabling more complex biological questions

Below is a recap of the main updates I could gather from the meeting.

📅 Next Tuesday I’ll publish a second article focused on the spatial biology updates from AGBT 2026.

Industrial-scale Single-cell Workflows 🏭

Roche’s new sequencer: AXELIOS 1

One of the most talked-about launches at the meeting was Roche’s new sequencer, AXELIOS 1, built around SBX (sequencing by expansion).

Why does this matter for single-cell?

Read more

It was “How fast can they scale, and how cleanly can they plug into clinical and translational workflows?”

Across announcements, 2 shifts stood out: single-cell is being treated as an industrial-grade data engine, and spatial biology is being positioned as a core layer of biological infrastructure rather than a niche imaging add-on.

Single-Cell Biology: from discovery tool to clinical trajectory

Single-cell was framed less as an emerging toolkit and more as a mature platform category powering translational research and drug discovery, especially as datasets move from “big” to “institution-scale.”

Read more