virtual spatial transcriptomics use case
Build a wider study from the samples you measured.
The slides you have already profiled can contribute to research across the rest of your cohort. With our team, you can use those measured examples to adapt DeepSpot-M to your material, then investigate predicted expression in samples that have not received a spatial assay.
Worked through for Sequencing & Core Facilities.
The research opportunity
How far could you take a research question by building on the samples you have already measured?
Why it matters
Carry a finding beyond the measured subset
A result from a few profiled samples can become a question about the rest of the cohort. You may want to investigate whether a predicted pattern recurs in other cases, or examine how it varies across their tissue. Adaptation lets you use measured examples from your own study before making predictions on the remaining slides.
Because those examples pair your H&E images with spatial measurements, they connect the analysis to the material you work with. The measured subset takes on a second role in supporting investigation across the collection.
Build on your laboratory's material
Your existing spatial data can support an analysis suited to the cohort you want to study. Fixation, staining and scanning shape the images a laboratory produces. Our team works with you to choose measured examples for adaptation, taking the material and scope of the intended study into account.
While the DeepSpot-M preprint demonstrates adaptation from one measured slide, the appropriate examples for your study depend on the cohort. After adaptation, you can investigate the remaining images on the same predicted gene panel.
Revisit questions left open by missing genes
Genes removed during quality control need not end every line of investigation involving them. In a separate restoration analysis, selected genes return as labelled predictions from the image. This lets you examine their predicted spatial patterns alongside the measurements retained from the experiment.
A question interrupted by missing values can then be considered again in its tissue context. You can assess the resulting hypothesis before committing material to another measurement.
Explore the possibilities
The DeepSpot-M preprint demonstrates adaptation using one measured slide and restoration of selected genes, extending what researchers can explore from an existing experiment.
Take existing measurements further
Adaptation extends investigation to other samples in the cohort. Restoration lets you revisit selected genes within an existing experiment, opening another path through the data you already hold.
Your input and output
- Input
- A few slides with measured spatial expression and a matching H&E image for each.
- Output
- Cohort-adapted virtual spatial transcriptomics and labelled predictions for restored genes.
Take the next step
Start with an example slide, or talk to us about your cohort.