Choosing the Right Spatial Biology Platform

What Is Spatial Biology?

Spatial biology is defined as the study of cells in their native tissue environment. Spatial biology as we know it today was born in 2019. It brought with it astounding new discoveries that have impacted both research and clinical pathology.

Spatial's First Tools

The first commercial platforms (GeoMx and Visium) utilized DNA-barcoded probes to capture transcriptomics information with an NGS sequencing readout. GeoMx also used barcoded antibodies to capture protein information with an nCounter or NGS readout. This first generation of spatial biology instruments could detect the whole transcriptomes from a group of cells, ~200 minimum, or protein from a group of ~20 cells.

Although these initial tools didn’t deliver single-cell resolution,  the discoveries that came from these instruments are world changing. Data from this new discipline changed the landscape of cancer research, Alzheimer's disease, and our understanding of the tumor microenvironment (TME). 

The Next Generation

The early systems did not offer single-cell resolution. That has changed in the second generation of instruments, including the Xenium, CosMx, PhenoImager, and MerFish. These instruments use cyclic fluorescent in situ hybridization (FISH) and function as imagers. 

The cyclic FISH process utilizes in situ hybridization with DNA barcoded probes followed by 4-color FISH probes that target the unique barcodes assigned to each transcript. The readout is a 16-bit barcode that is decoded on the instrument resulting in the transcript counts per cell output. All of these imagers provide true single-cell resolution with x, y, z, coordinates and cell boundaries. Data quality is excellent based on our experience, but instrument run times can be long depending on the plex of the assay; CosMx highest plex assay is 6K RNA and Xenium’s highest plex is 5K RNA.

Depiction of the cyclic FISH process that generates the 16-bit barcode.

The Newest Spatial Tools

Visium HD, CosMx 6K RNA, and Xenium 5K are excellent platforms to identify targets and pathways involved in research questions. Visium HD is unique in that it delivers a whole transcriptome NGS readout at 2µM resolution on FFPE tissues, but note that the data only include x,y coordinates for the transcripts.The recent launch of Visium HD by 10x Genomics is truly revolutionary for any scientist looking to do discovery. 

Following the discovery runs, for the next set of experiments it is logical to step down in plex so that the focus is on the key pathways already identified. As a bonus, the cost savings are often significant with this approach.  

Getting Started: 7 Key Questions to Choose the Right Spatial Platform

Spatial biology is the hot-topic technology of the year — so much so that Nature Genetics issued a call for spatial omics submissions in January of 2024. It's understandable that you may be considering adding spatial biology to your current project. But what kind of spatial biology is best for the project (and most likely to get the paper into Nature Genetics)?

The remainder of this article covers a series of seven important questions that will clarify the type of spatial biology that will be most beneficial to the needs of your project.

1. What is the purpose of the spatial experiment?

Is the goal to identify novel biomarkers? Or is it to monitor a series of known pathways for the impact of a treatment regimen?

These two questions can help you determine if the project is at the early discovery stage or at a more focused point. Discovery projects typically aim for the highest plex available, like the Visium HD whole transcriptome, CosMx 6k RNA, or Xenium 5k RNA. Projects with more limited scopes can choose biologically focused panels that offer lower plex than the discovery platforms.

Maybe your project aims to do both. In this case, many researchers have found success by first running discovery on a small number of samples. Then, they move on to a lower plex focused panel that delivers the specific information to advance their research. The smaller panels are typically less costly and have shorter turn-around times.

2. What level of resolution is needed to address the experimental question?

True single-cell resolution is achieved with imaging-based platforms like CosMx and Xenium. These instruments output RNA transcripts in x, y, and z coordinates.

RNA transcripts with x, y coordinates are generated by transcriptomics platforms such as Visium HD and Curio Trekker. These instruments provide sequencing readouts, but no subcellular imaging component.

3. What species is being assayed?

Some platforms only support certain species with off-the-shelf panels, usually human and mouse. It’s important to find out early if the technology in question has pre-designed panels for the organism being studied. If not, some platforms support custom designs for certain species.

Examples of off-the-shelf panels limited to Human and Mouse for Xenium and CosMx spatial biology platforms.

4. How are the tissue samples prepared?

Are the samples stored as fresh frozen tissue (FF), or are they formalin-fixed paraffin embedded (FFPE)? If the samples are collected in-house, it is possible to choose the method of preparation. However, if samples are coming from a biobank, options may be more limited. Some platforms are only validated for either FFPE or fresh frozen tissue, not both.

List of spatial biology instruments with their unique sample preparation protocols.

5. What kind of data analysis tools are available?

Often the most cutting-edge platforms are launched with rudimentary analysis tools that offer little more than QC. Even if the platform does provide some tools for analysis, the visualizations and figures might not be publication quality. In this age of multiomics, it is not enough to merely utilize orthogonal methods — it is now essential to combine various modes of data together for a comprehensive understanding of the subject.

6. In terms of experimental design, how many control and experimental samples will be included?

Biostatisticians require a minimum of three replicates to conduct statistically relevant calculations. However, replicates in spatial biology are often limited to an n=2, with full acceptance from the majority of reviewers.

To put it bluntly, spatial biology experiments are expensive. That limits everyone’s ability to run replicates. A common approach to incorporate biostatistics is that regions of interest are often selected within each tissue and used for the comparisons.

Differences between CosMx, Xenium, and Visium HD slides.

7. How many sections of each type can you fit onto each slide for the platform selected?

Every spatial biology platform has a strict area on the slide where the tissue must be placed. In some cases, if tissue or paraffin exceeds that area, the slide cannot be processed. In other cases, tissue that extends outside of the demarcated will not be analyzed and is typically destroyed during slide processing.

In addition, spatial experiments are prone to bias. It is extremely important to have a control and an experimental section on the same slide. It is possible to address this variance downstream with linear mixed models, but from an experimental design perspective, it’s always best to minimize variability whenever possible.

 Psomagen’s team has years of spatial biology expertise and is here to help design the experiment that addresses your project’s questions at a reasonable cost. Any spatial biology project starts with good experimental design — because this list isn’t comprehensive, please contact us so our spatial biology experts can help you meet your research goals.