[moved] How to align image to cropped data, visualize with Xenium Explorer

[LucaMarconato]

Moved from GitHub Coordinates from Xenium Explorer not matching the original data · Issue #167 · scverse/spatialdata-io · GitHub, originally asked from Yifan-debug (yifan wang) · GitHub.

Original question: in my case, I only focus on a small region of the whole slide. So I only take high-resolution confocal images specifically of this small region after Xenium. I’m wondering if there is any tutorial I could follow not only for cropping the image but also for aligning my confocal image to the cropped image and outputting a folder that the locations of transcripts can be visualized in Xenium Explorer.

[LucaMarconato]

Here is the answer.
We don’t have a tutorial for what you asked exactly, but the various compoents to achieve this are explained in some other tutorials.

• Aligning the image:
  • you can choose to align the image before cropping, or to crop the data and then align the image to the cropped data; as you prefer. The tutorial to align images (a landmark-based approach is used) is shown here: Use landmark annotations to align multiple -omics layers — spatialdata
  • Alternatively you can align the data with Xenium Explorer and export the transformation as shown in the notebook spatialdata_xenium_explorer/docs/10x_tutorials/xenium/xenium_tuto.ipynb at master · quentinblampey/spatialdata_xenium_explorer · GitHub.
• Saving the Xenium cropped data to visualize it with Xenium explorer: the same notebook shows how to export any Xenium SpatialData object to visualize it with Xenium explorer. Here is the notebook again:

Hope this leads to the solution, let me know if you have any other question.