Thank you for such a well-crafted and insightful piece, this is a rare example of a discussion that is both detailed and grounded in the realities of the field. I particularly appreciate the acknowledgment of what has already been achieved in spatial transcriptomics, which is critical context as new technologies are introduced.
To build on your perspective, a few additional points may be worth considering:
- First, whole transcriptome spatial biology at true single-cell resolution is no longer aspirational.
With the introduction of the CosMx® Whole Transcriptome assay by Bruker Spatial Biology at AACR 2025, imaging-based platforms have already demonstrated subcellular, spatial whole transcriptome profiling directly from FFPE tissues—arguably one of the most meaningful inflection points for translational research.
- Second, adoption and validation are already underway.
What was once considered technically out of reach is now being applied broadly, with a growing body of preprints and peer-reviewed studies using CosMx whole transcriptome assay (including work published in the Journal of Clinical Investigation https://pmc.ncbi.nlm.nih.gov/articles/PMC12435852/) demonstrating real-world utility across disease areas.
- Third, scale should be viewed in context—not isolation.
Throughput is important, but it is only one dimension. In practice, researchers routinely scale studies through parallelization (e.g., multiple instruments), allowing them to preserve sensitivity and resolution rather than trade them off for speed.
Recent comparative work from Northwestern University (https://www.nature.com/articles/s41467-026-72482-w) highlights a critical point: detection sensitivity and gene diversity directly impact discovery. CosMx’s direct hybridization chemistry achieved higher transcript detection and broader gene diversity compared to RCA-based approaches like Xenium. This increased sensitivity enabled detection of biologically critical signals—such as FOXP3+ regulatory T cells, CSF3, and PGK1—that were not resolved by Xenium under comparable conditions.
This raises an important consideration for emerging platforms like Atera, which are also built on RCA chemistry: while whole transcriptome scale is promising, real-world sensitivity and biological resolution remain key benchmarks yet to be demonstrated.
- Fifth, robustness in real-world FFPE samples is non-negotiable.
FFPE compatibility is essential, but performance across variable tissue quality, processing conditions, and multi-site studies is what ultimately determines utility at scale. Comparative analyses (e.g., work by Dr. Dulai and colleagues https://www.nature.com/articles/s41467-026-72482-w) have shown CosMx to maintain consistent performance under these conditions, whereas RCA-based approaches required additional pre-screening steps, introducing complexity that may limit scalability in practice. Atera rely on the same foundational RCA chemistry as in Xenium and is unproven in this critical context.
- Finally, multiomic context remains a major differentiator.
Transcriptomic data alone provides only part of the picture. The integration of protein-level information—capturing functional and post-translational states—adds a critical layer of biological insight that is increasingly important for confident interpretation. At present, the Atera platform lacks protein capabilities and CosMx is the only commercially available platform capable of imaging high-plex protein with whole transcriptome on the same slide with single cell subcellular resolution.
Taken together, these considerations reinforce that platform evaluation is inherently multidimensional, extending beyond scale to include sensitivity, robustness, and depth of biological insight.
Appreciate you advancing this conversation in such a thoughtful way, this level of discourse is exactly what the field needs right now.
My politically INcorrect impressions after the webinar, in addition to what you already point out in your article:
- Atera seems very much Xenium with Flex probes (is it only me?), with 4 slides/run instead of 2...the instrument that will be used for "decades" to come. Not sure all the cores/universities/institutes that invested LOTS of money to buy a Xenium will be happy to hear that (an instrument that was also sold as "revolutionary" and "groundbreaking"...)
- the base-by-base sequencing they sold as "groundbreaking" is on ElemBio's roadmap since a year at least
- I also thought the same: Xenium with Flex probes. And I also thought...if we are using the same probes for the Flex, Xenium, and Atera, are we really validating results? Or are we just repeating the same measurement in a different "instrument"?
- I also expect that people that invested recently in Xenium might not be very happy.
- Thanks for reminding me of the Element Bio AVITI24 roadmap. Yes, they mention that it will be "Available in H2 2026."
Thank you for such a well-crafted and insightful piece, this is a rare example of a discussion that is both detailed and grounded in the realities of the field. I particularly appreciate the acknowledgment of what has already been achieved in spatial transcriptomics, which is critical context as new technologies are introduced.
To build on your perspective, a few additional points may be worth considering:
- First, whole transcriptome spatial biology at true single-cell resolution is no longer aspirational.
With the introduction of the CosMx® Whole Transcriptome assay by Bruker Spatial Biology at AACR 2025, imaging-based platforms have already demonstrated subcellular, spatial whole transcriptome profiling directly from FFPE tissues—arguably one of the most meaningful inflection points for translational research.
- Second, adoption and validation are already underway.
What was once considered technically out of reach is now being applied broadly, with a growing body of preprints and peer-reviewed studies using CosMx whole transcriptome assay (including work published in the Journal of Clinical Investigation https://pmc.ncbi.nlm.nih.gov/articles/PMC12435852/) demonstrating real-world utility across disease areas.
- Third, scale should be viewed in context—not isolation.
Throughput is important, but it is only one dimension. In practice, researchers routinely scale studies through parallelization (e.g., multiple instruments), allowing them to preserve sensitivity and resolution rather than trade them off for speed.
- Fourth, sensitivity ultimately defines biological insight.
Recent comparative work from Northwestern University (https://www.nature.com/articles/s41467-026-72482-w) highlights a critical point: detection sensitivity and gene diversity directly impact discovery. CosMx’s direct hybridization chemistry achieved higher transcript detection and broader gene diversity compared to RCA-based approaches like Xenium. This increased sensitivity enabled detection of biologically critical signals—such as FOXP3+ regulatory T cells, CSF3, and PGK1—that were not resolved by Xenium under comparable conditions.
This raises an important consideration for emerging platforms like Atera, which are also built on RCA chemistry: while whole transcriptome scale is promising, real-world sensitivity and biological resolution remain key benchmarks yet to be demonstrated.
- Fifth, robustness in real-world FFPE samples is non-negotiable.
FFPE compatibility is essential, but performance across variable tissue quality, processing conditions, and multi-site studies is what ultimately determines utility at scale. Comparative analyses (e.g., work by Dr. Dulai and colleagues https://www.nature.com/articles/s41467-026-72482-w) have shown CosMx to maintain consistent performance under these conditions, whereas RCA-based approaches required additional pre-screening steps, introducing complexity that may limit scalability in practice. Atera rely on the same foundational RCA chemistry as in Xenium and is unproven in this critical context.
- Finally, multiomic context remains a major differentiator.
Transcriptomic data alone provides only part of the picture. The integration of protein-level information—capturing functional and post-translational states—adds a critical layer of biological insight that is increasingly important for confident interpretation. At present, the Atera platform lacks protein capabilities and CosMx is the only commercially available platform capable of imaging high-plex protein with whole transcriptome on the same slide with single cell subcellular resolution.
Taken together, these considerations reinforce that platform evaluation is inherently multidimensional, extending beyond scale to include sensitivity, robustness, and depth of biological insight.
Appreciate you advancing this conversation in such a thoughtful way, this level of discourse is exactly what the field needs right now.
Thank you SO MUCH for this comment. It really adds a lot of information. Things that I didn’t know!
My politically INcorrect impressions after the webinar, in addition to what you already point out in your article:
- Atera seems very much Xenium with Flex probes (is it only me?), with 4 slides/run instead of 2...the instrument that will be used for "decades" to come. Not sure all the cores/universities/institutes that invested LOTS of money to buy a Xenium will be happy to hear that (an instrument that was also sold as "revolutionary" and "groundbreaking"...)
- the base-by-base sequencing they sold as "groundbreaking" is on ElemBio's roadmap since a year at least
- I also thought the same: Xenium with Flex probes. And I also thought...if we are using the same probes for the Flex, Xenium, and Atera, are we really validating results? Or are we just repeating the same measurement in a different "instrument"?
- I also expect that people that invested recently in Xenium might not be very happy.
- Thanks for reminding me of the Element Bio AVITI24 roadmap. Yes, they mention that it will be "Available in H2 2026."