Achieving genuine spatial biology automation is what actually separates a technique that produces a handful of publishable images from one that can support a real pharma pipeline processing hundreds of samples across an entire drug discovery program. The underlying biology has been ready for years; the throughput math has not, and closing that gap is now where much of the field’s practical innovation is concentrated.
Key takeaways
|
Why throughput is spatial’s bottleneck
A spatial biology experiment performed manually involves a long sequence of precise, sequential steps, tissue sectioning, fixation or antigen retrieval, staining or probe hybridization, imaging and downstream analysis, each requiring careful technique and each vulnerable to operator-to-operator variability if performed by hand. That sequence is entirely tractable for a handful of research samples. It becomes the limiting factor the moment a program needs data from hundreds of samples across a drug discovery pipeline, since no realistic number of skilled technicians can execute that sequence manually at pharma scale without introducing exactly the kind of variability spatial biomarker validation, covered elsewhere in this cluster, is specifically designed to catch.
Our earlier coverage in The future of spatial biology depends on automation establishes the general case for why automation matters to this field. This guide goes further, into the specific throughput figures that quantify what industrial-scale spatial sample processing actually looks like today.
Automating sample prep and imaging
The clearest concrete demonstration of what automated sample preparation can achieve at scale comes from deep visual proteomics, a spatial proteomics method combining whole-slide imaging, machine-learning-guided image analysis, single-cell laser microdissection, and ultrasensitive mass spectrometry.
Hundreds of samples a day, not hours per sampleTo automate tissue processing following laser microdissection, researchers developed a robotic sample processing protocol built around an automated single-cell dispensing platform, enabling the preparation of hundreds of tissue samples per day. That capability supports an ultrasensitive spatial tissue proteomics workflow described as exceeding 100 proteomes per day, a figure that reframes what "high throughput" means for a technique that not long ago required painstaking manual handling of each individual microdissected sample. Reaching that throughput required automating the specific step that had been the practical bottleneck, sample handling immediately after microdissection, rather than automating the whole workflow indiscriminately. That is a generalizable lesson: identifying which single step actually limits daily sample count, and automating that step specifically, is usually more tractable than attempting to automate an entire multi-step protocol simultaneously. |
A second example demonstrates the same principle using standard, already-widespread laboratory automation rather than bespoke spatial-specific hardware. Researchers automated the library construction step of a widely used spatial transcriptomics protocol using the Agilent Bravo Liquid Handling Platform, a robotic workstation already common across genomics laboratories generally. This approach increased throughput and robustness of library construction while reducing hands-on time relative to the manual protocol, and doing so on a platform many labs may already own for other genomics applications lowers the practical adoption barrier considerably compared with a purpose-built, spatial-specific automation system.
Data pipeline automation
Physical sample processing is only half of the throughput problem. The computational side, turning raw images or sequencing output into analyzed, interpretable results, faces its own scaling challenge, and it is a distinct question this cluster addresses in depth elsewhere.
Automating the computational discovery of spatial biomarker candidates specifically, including interpretable feature extraction frameworks and automated candidate-ranking methods, is covered in full in Spatial Biology in Biomarker Discovery. The throughput principle established there, replacing manual visual inspection with computational search because data volume makes manual review impractical, is the same underlying logic driving the physical automation described in this guide; the two simply apply that logic to different stages of the same overall pipeline.
Standardization at scale
Automating a single site’s workflow and standardizing that automated workflow across multiple sites are genuinely different achievements, and the harder of the two remains less solved.
A 2025 study addresses this from a cost and accessibility angle specifically. Researchers introduced PRISMS, an open-sourced, automated multiplexing pipeline using a liquid handling robot with thermal control for rapid, automated staining of RNA and protein samples, compatible with multiple biospecimen types and streamlined microscopy software. The motivation is stated directly: high costs associated with proprietary instrumentation, specialized reagents, and complex workflows have limited broad application of spatial omics techniques, and an open-source approach addresses that limitation by design rather than by discounting an existing proprietary system.
That distinction matters for standardization specifically, since a proprietary automated platform standardizes a workflow only among the labs that can afford it, while an open, documented protocol on widely available hardware has a more realistic path toward broad, cross-site standardization, precisely because more sites can actually adopt it in the first place. Standardization that only the best-funded labs can reach is standardization in a narrower and less useful sense than the field ultimately needs.
The industrialization roadmap
Three practical steps recur across the automation efforts described in this guide, worth treating as a general sequence rather than a single leap to full automation.
- Identify the specific bottleneck step first. The deep visual proteomics example above succeeded by automating the single step, post-microdissection sample handling, that actually limited daily throughput, rather than attempting to automate an entire workflow uniformly from the outset.
- Prefer standard, already-adopted platforms where feasible. Building automation on hardware, such as the Agilent Bravo platform, that a lab may already operate for other genomics work lowers the adoption barrier relative to bespoke, spatial-specific instrumentation.
- Treat cross-site standardization as its own distinct milestone, not an automatic byproduct of automating one site. An automated protocol that works reliably at the site that built it has not yet demonstrated it will perform identically elsewhere; that cross-site consistency question connects directly to the harmonization evidence covered in this cluster’s biomarker development guides.
Broader questions of how a scaled, automated spatial workflow fits into a regulated laboratory’s operational and workflow standards are addressed by our colleagues at Lab Manager in Running Spatial Biology in the Lab: Workflow, Throughput, and Sample Management.
For the broader convergence argument and clinical infrastructure this spoke builds on, see Translating Spatial Biology into the Clinic: Digital Pathology and Beyond, and for where this fits within the full spatial biology pipeline, Spatial Biology in Drug Discovery: From Target Discovery to Translational Medicine.
This article was produced under Drug Discovery News’s AI editorial policies.
