Skip to content

Is Aurora right for your research?

Start with your research question

Aurora predicts spatial gene expression from routine H&E images, helping researchers explore molecular patterns across samples, cohorts and disease.

Aurora works best when

Your slidesH&E slides you already hold: an archive, a cohort or the sections of a study.
Your questionWhere molecular patterns appear across the tissue, and how they differ between regions or samples.
What you getPredicted expression for every gene the model returns, at every tile kept, on your slide's own coordinates.

Common applications

  • Explore existing collections. Ask new questions of archived tissue and cohorts.
  • Prioritise follow-up research. Choose samples and regions for the next experiment.
  • Build on molecular data. Add sample-matched bulk RNA to the spatial prediction.

Explore all research applications.

Before you start

Bring a readable H&E scan and a research question. If the scan is faint or blurred, check a small region first. Choose a workflow according to whether your institution permits image upload.

Virtual spatial transcriptomics is predicted, spatially resolved gene expression from H&E images. Aurora returns it on the slide's own coordinates as one transcriptome-wide matrix. No spatial assay is run.

Where your images go

Choose where image processing happens. Molecular prediction runs on Aurora infrastructure with either workflow.

DeepSpot-H turns each tile of your slide into one embedding. It can run in your environment or on Aurora infrastructure. DeepSpot-M turns those embeddings into spatial gene expression. It runs on Aurora infrastructure.

Send your slides Your slides stay with you
Choose this when Your images may leave your machine. Your images may not leave your machine.
What you need from us A key, and nothing else. A key. The weights for the local step download once, onto your machine.
What you send Tiles cut from your slide, in one archive. Numbers computed from your tiles, with their positions.
What leaves your machine The tile images. No image.
Continue Send your slides Your slides stay with you

Which model reads it

The tutorial examples select a published model that accepts your prepared input. The API catalogue lists release identifiers and accepted inputs.

Where to go next

Both routes begin by cutting and checking tiles on your machine. Continue with Install and prepare a slide, or explore one real slide first.